Added fix for loading C++ transformations, reversed logic whether result has been added

.clang-tidy

@@ -2,6 +2,7 @@
 Checks: '*,
          -abseil-string-find-str-contains,
          -altera-struct-pack-align,
+         -altera-unroll-loops,
          -android-*,
          -cert-err58-cpp,
          -cppcoreguidelines-avoid-c-arrays,
@@ -58,7 +59,8 @@ Checks: '*,
          -readability-implicit-bool-conversion,
          -readability-magic-numbers,
          -readability-named-parameter,
-         -misc-no-recursion'
+         -misc-no-recursion,
+         -concurrency-mt-unsafe'
 WarningsAsErrors: ''
 HeaderFilterRegex: 'src/.*'
 AnalyzeTemporaryDtors: false

.githooks/pre-commit

@@ -32,7 +32,7 @@ for file in $modified_files; do
   fi
 
   echo "Running header checker..."
-  $project_folder/tools/header-checker.py $tmpdir/$file
+  $project_folder/tools/header-checker.py $tmpdir/$file $file --amend-year
   code=$?
 
   if [ $code -ne 0 ]; then

.github/workflows/daily_benchmark.yaml

@@ -25,7 +25,7 @@ jobs:
       - name: Build release binaries
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Initialize dependencies.
           ./init

.github/workflows/diff.yaml

@@ -32,7 +32,7 @@ jobs:
       - name: Build community binaries
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Initialize dependencies.
           ./init
@@ -45,7 +45,7 @@ jobs:
       - name: Run unit tests
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run unit tests.
           cd build
@@ -70,7 +70,7 @@ jobs:
       - name: Build combined ASAN, UBSAN and coverage binaries
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Initialize dependencies.
           ./init
@@ -82,7 +82,7 @@ jobs:
       - name: Run unit tests
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run unit tests. It is restricted to 2 threads intentionally, because higher concurrency makes the timing related tests unstable.
           cd build
@@ -91,7 +91,7 @@ jobs:
       - name: Compute code coverage
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Compute code coverage.
           cd tools/github
@@ -109,7 +109,7 @@ jobs:
 
       - name: Run clang-tidy
         run: |
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Restrict clang-tidy results only to the modified parts
           git diff -U0 master... -- src ':!*.hpp' | ./tools/github/clang-tidy/clang-tidy-diff.py -p 1 -j $THREADS -path build | tee ./build/clang_tidy_output.txt
@@ -136,7 +136,7 @@ jobs:
       - name: Build debug binaries
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Initialize dependencies.
           ./init
@@ -149,7 +149,7 @@ jobs:
       - name: Run leftover CTest tests
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run leftover CTest tests (all except unit and benchmark tests).
           cd build
@@ -181,7 +181,7 @@ jobs:
       - name: Run cppcheck and clang-format
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run cppcheck and clang-format.
           cd tools/github
@@ -212,7 +212,7 @@ jobs:
       - name: Build release binaries
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Initialize dependencies.
           ./init
@@ -238,7 +238,7 @@ jobs:
       - name: Run unit tests
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run unit tests.
           cd build
@@ -272,7 +272,7 @@ jobs:
       - name: Create enterprise DEB package
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           cd build
 
@@ -319,7 +319,7 @@ jobs:
       - name: Build release binaries
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Initialize dependencies.
           ./init
@@ -360,7 +360,7 @@ jobs:
       - name: Build release binaries
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Initialize dependencies.
           ./init

.github/workflows/full_clang_tidy.yaml

@@ -23,7 +23,7 @@ jobs:
       - name: Build debug binaries
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Initialize dependencies.
           ./init
@@ -36,10 +36,10 @@ jobs:
 
       - name: Run clang-tidy
         run: |
-          source /opt/toolchain-v3/activate
+          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 $THREADS -clang-tidy-binary=/opt/toolchain-v3/bin/clang-tidy "$PWD/src/*" |
+          ./tools/github/clang-tidy/run-clang-tidy.py -p build -j $THREADS -clang-tidy-binary=/opt/toolchain-v4/bin/clang-tidy "$PWD/src/*" |
             tee ./build/full_clang_tidy_output.txt
 
       - name: Summarize clang-tidy results

.github/workflows/package_all.yaml

@@ -101,7 +101,7 @@ jobs:
       - name: "Build package"
         run: |
           cd release/package
-          ./run.sh package debian-10 --for-docker
+          ./run.sh package debian-11 --for-docker
           ./run.sh docker
       - name: "Upload package"
         uses: actions/upload-artifact@v2
@@ -143,7 +143,7 @@ jobs:
           name: ubuntu-2004
           path: build/output/ubuntu-20.04/memgraph*.deb
 
-  debian-10-platform:
+  debian-11-platform:
     runs-on: [self-hosted, DockerMgBuild]
     timeout-minutes: 60
     steps:
@@ -153,9 +153,9 @@ jobs:
           fetch-depth: 0 # Required because of release/get_version.py
       - name: "Build package"
         run: |
-          ./release/package/run.sh package debian-10 --for-platform
+          ./release/package/run.sh package debian-11 --for-platform
       - name: "Upload package"
         uses: actions/upload-artifact@v2
         with:
-          name: debian-10-platform
-          path: build/output/debian-10/memgraph*.deb
+          name: debian-11-platform
+          path: build/output/debian-11/memgraph*.deb

.github/workflows/release_centos8.yaml

@@ -26,7 +26,7 @@ jobs:
       - name: Build community binaries
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Initialize dependencies.
           ./init
@@ -39,7 +39,7 @@ jobs:
       - name: Run unit tests
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run unit tests.
           cd build
@@ -64,7 +64,7 @@ jobs:
       - name: Build coverage binaries
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Initialize dependencies.
           ./init
@@ -77,7 +77,7 @@ jobs:
       - name: Run unit tests
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run unit tests.
           cd build
@@ -86,7 +86,7 @@ jobs:
       - name: Compute code coverage
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Compute code coverage.
           cd tools/github
@@ -121,7 +121,7 @@ jobs:
       - name: Build debug binaries
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Initialize dependencies.
           ./init
@@ -134,7 +134,7 @@ jobs:
       - name: Run leftover CTest tests
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run leftover CTest tests (all except unit and benchmark tests).
           cd build
@@ -166,7 +166,7 @@ jobs:
       - name: Run cppcheck and clang-format
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run cppcheck and clang-format.
           cd tools/github
@@ -198,7 +198,7 @@ jobs:
       - name: Build release binaries
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Initialize dependencies.
           ./init
@@ -211,7 +211,7 @@ jobs:
       - name: Create enterprise RPM package
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           cd build
 
@@ -233,7 +233,7 @@ jobs:
       - name: Run micro benchmark tests
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run micro benchmark tests.
           cd build
@@ -272,7 +272,7 @@ jobs:
       - name: Run unit tests
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run unit tests.
           cd build

.github/workflows/release_debian10.yaml

@@ -26,7 +26,7 @@ jobs:
       - name: Build community binaries
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Initialize dependencies.
           ./init
@@ -39,7 +39,7 @@ jobs:
       - name: Run unit tests
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run unit tests.
           cd build
@@ -64,7 +64,7 @@ jobs:
       - name: Build coverage binaries
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Initialize dependencies.
           ./init
@@ -77,7 +77,7 @@ jobs:
       - name: Run unit tests
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run unit tests.
           cd build
@@ -86,7 +86,7 @@ jobs:
       - name: Compute code coverage
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Compute code coverage.
           cd tools/github
@@ -121,7 +121,7 @@ jobs:
       - name: Build debug binaries
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Initialize dependencies.
           ./init
@@ -134,7 +134,7 @@ jobs:
       - name: Run leftover CTest tests
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run leftover CTest tests (all except unit and benchmark tests).
           cd build
@@ -166,7 +166,7 @@ jobs:
       - name: Run cppcheck and clang-format
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run cppcheck and clang-format.
           cd tools/github
@@ -198,7 +198,7 @@ jobs:
       - name: Build release binaries
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Initialize dependencies.
           ./init
@@ -211,7 +211,7 @@ jobs:
       - name: Create enterprise DEB package
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           cd build
 
@@ -232,7 +232,7 @@ jobs:
       - name: Run micro benchmark tests
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run micro benchmark tests.
           cd build
@@ -271,7 +271,7 @@ jobs:
       - name: Run unit tests
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run unit tests.
           cd build
@@ -333,7 +333,7 @@ jobs:
       - name: Build release binaries
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Initialize dependencies.
           ./init

.github/workflows/release_ubuntu2004.yaml

@@ -26,7 +26,7 @@ jobs:
       - name: Build community binaries
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Initialize dependencies.
           ./init
@@ -39,7 +39,7 @@ jobs:
       - name: Run unit tests
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run unit tests.
           cd build
@@ -64,7 +64,7 @@ jobs:
       - name: Build coverage binaries
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Initialize dependencies.
           ./init
@@ -77,7 +77,7 @@ jobs:
       - name: Run unit tests
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run unit tests.
           cd build
@@ -86,7 +86,7 @@ jobs:
       - name: Compute code coverage
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Compute code coverage.
           cd tools/github
@@ -121,7 +121,7 @@ jobs:
       - name: Build debug binaries
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Initialize dependencies.
           ./init
@@ -134,7 +134,7 @@ jobs:
       - name: Run leftover CTest tests
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run leftover CTest tests (all except unit and benchmark tests).
           cd build
@@ -166,7 +166,7 @@ jobs:
       - name: Run cppcheck and clang-format
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run cppcheck and clang-format.
           cd tools/github
@@ -198,7 +198,7 @@ jobs:
       - name: Build release binaries
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Initialize dependencies.
           ./init
@@ -211,7 +211,7 @@ jobs:
       - name: Create enterprise DEB package
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           cd build
 
@@ -232,7 +232,7 @@ jobs:
       - name: Run micro benchmark tests
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run micro benchmark tests.
           cd build
@@ -271,7 +271,7 @@ jobs:
       - name: Run unit tests
         run: |
           # Activate toolchain.
-          source /opt/toolchain-v3/activate
+          source /opt/toolchain-v4/activate
 
           # Run unit tests.
           cd build

CMakeLists.txt

@@ -322,7 +322,8 @@ if (UBSAN)
   # To make the program abort on undefined behavior, use UBSAN_OPTIONS=halt_on_error=1.
 endif()
 
-set(MG_PYTHON_VERSION "" CACHE STRING "Specify the exact python version used by the query modules")
+set(MG_PYTHON_VERSION "" CACHE STRING "Specify the exact Python version used by the query modules")
+set(MG_PYTHON_PATH "" CACHE STRING "Specify the exact Python path used by the query modules")
 
 # Add subprojects
 include_directories(src)

README.md

@@ -115,6 +115,14 @@ understand what actions will and will not be tolerated.
 Read our [contributing guide](CONTRIBUTING.md) to learn about our development
 process and how to propose bug fixes and improvements.
 
+### Internals
+
+Read our
+[internal](https://memgraph.notion.site/Memgraph-Internals-12b69132d67a417898972927d6870bd2)
+docs to learn more about Memgraph's architecture, how to build the project from
+source and how to start contributing. All information related to the database,
+can be found in the aforementioned docs.
+
 ### :scroll: License
 
 Memgraph Community is available under the [BSL

cmake/FindJemalloc.cmake

@@ -0,0 +1,55 @@
+# Try to find jemalloc library
+#
+# Use this module as:
+#   find_package(Jemalloc)
+#
+# or:
+#   find_package(Jemalloc REQUIRED)
+#
+# This will define the following variables:
+#
+#   Jemalloc_FOUND           True if the system has the jemalloc library.
+#   Jemalloc_INCLUDE_DIRS    Include directories needed to use jemalloc.
+#   Jemalloc_LIBRARIES       Libraries needed to link to jemalloc.
+#
+# The following cache variables may also be set:
+#
+# Jemalloc_INCLUDE_DIR     The directory containing jemalloc/jemalloc.h.
+# Jemalloc_LIBRARY         The path to the jemalloc static library.
+
+find_path(Jemalloc_INCLUDE_DIR NAMES jemalloc/jemalloc.h PATH_SUFFIXES include)
+
+find_library(Jemalloc_LIBRARY NAMES libjemalloc.a PATH_SUFFIXES lib)
+
+include(FindPackageHandleStandardArgs)
+find_package_handle_standard_args(Jemalloc
+  FOUND_VAR Jemalloc_FOUND
+  REQUIRED_VARS
+    Jemalloc_LIBRARY
+    Jemalloc_INCLUDE_DIR
+)
+
+if(Jemalloc_FOUND)
+  set(Jemalloc_LIBRARIES ${Jemalloc_LIBRARY})
+  set(Jemalloc_INCLUDE_DIRS ${Jemalloc_INCLUDE_DIR})
+else()
+  if(Jemalloc_FIND_REQUIRED)
+    message(FATAL_ERROR "Cannot find jemalloc!")
+  else()
+    message(WARNING "jemalloc is not found!")
+  endif()
+endif()
+
+if(Jemalloc_FOUND AND NOT TARGET Jemalloc::Jemalloc)
+  add_library(Jemalloc::Jemalloc UNKNOWN IMPORTED)
+  set_target_properties(Jemalloc::Jemalloc
+    PROPERTIES
+      IMPORTED_LOCATION "${Jemalloc_LIBRARY}"
+      INTERFACE_INCLUDE_DIRECTORIES "${Jemalloc_INCLUDE_DIR}"
+  )
+endif()
+
+mark_as_advanced(
+  Jemalloc_INCLUDE_DIR
+  Jemalloc_LIBRARY
+)

environment/README.md

@@ -1,4 +0,0 @@
-# Memgraph Build and Run Environments
-
-Please continue in
-[Notion](https://www.notion.so/memgraph/Tools-05e0baafb78a49b386e0063b4833d23d).

environment/os/centos-7.sh

@@ -14,9 +14,14 @@ TOOLCHAIN_BUILD_DEPS=(
     expat-devel libipt-devel libbabeltrace-devel xz-devel python3-devel # gdb
     texinfo # gdb
     libcurl-devel # cmake
+    curl # snappy
     readline-devel # cmake and llvm
     libffi-devel libxml2-devel perl-Digest-MD5 # llvm
     libedit-devel pcre-devel automake bison # swig
+    file
+    openssl-devel
+    gmp-devel
+    gperf
 )
 
 TOOLCHAIN_RUN_DEPS=(
@@ -26,10 +31,10 @@ TOOLCHAIN_RUN_DEPS=(
     expat libipt libbabeltrace xz-libs python3 # for gdb
     readline # for cmake and llvm
     libffi libxml2 # for llvm
+    openssl-devel
 )
 
 MEMGRAPH_BUILD_DEPS=(
-    git # source code control
     make pkgconfig # build system
     curl wget # for downloading libs
     libuuid-devel java-11-openjdk # required by antlr
@@ -49,6 +54,7 @@ MEMGRAPH_BUILD_DEPS=(
     doxygen graphviz # source documentation generators
     which mono-complete dotnet-sdk-3.1 golang nodejs zip unzip java-11-openjdk-devel # for driver tests
     autoconf # for jemalloc code generation
+    libtool  # for protobuf code generation
 )
 
 list() {
@@ -94,8 +100,12 @@ install() {
         echo "NOTE: export LANG=en_US.utf8"
     fi
     yum install -y epel-release
+    yum remove -y ius-release
+    yum install -y \
+      https://repo.ius.io/ius-release-el7.rpm
     yum update -y
-    yum install -y wget git python3 python3-pip
+    yum install -y wget python3 python3-pip
+    yum install -y git224
     for pkg in $1; do
         if [ "$pkg" == libipt ]; then
             if ! yum list installed libipt >/dev/null 2>/dev/null; then

environment/os/centos-8.sh

@@ -13,9 +13,14 @@ TOOLCHAIN_BUILD_DEPS=(
     zlib-devel # zlib library used for all builds
     expat-devel libipt-devel libbabeltrace-devel xz-devel python36-devel texinfo # for gdb
     libcurl-devel # for cmake
+    curl # snappy
     readline-devel # for cmake and llvm
     libffi-devel libxml2-devel # for llvm
     libedit-devel pcre-devel automake bison # for swig
+    file
+    openssl-devel
+    gmp-devel
+    gperf
 )
 
 TOOLCHAIN_RUN_DEPS=(
@@ -25,6 +30,7 @@ TOOLCHAIN_RUN_DEPS=(
     expat libipt libbabeltrace xz-libs python36 # for gdb
     readline # for cmake and llvm
     libffi libxml2 # for llvm
+    openssl-devel
 )
 
 MEMGRAPH_BUILD_DEPS=(
@@ -48,6 +54,7 @@ MEMGRAPH_BUILD_DEPS=(
     which mono-complete dotnet-sdk-3.1 nodejs golang zip unzip java-11-openjdk-devel # for driver tests
     sbcl # for custom Lisp C++ preprocessing
     autoconf # for jemalloc code generation
+    libtool  # for protobuf code generation
 )
 
 list() {

environment/os/debian-10.sh

@@ -15,6 +15,12 @@ TOOLCHAIN_BUILD_DEPS=(
     libcurl4-openssl-dev # for cmake
     libreadline-dev # for cmake and llvm
     libffi-dev libxml2-dev # for llvm
+    curl # snappy
+    file # for libunwind
+    libssl-dev # for libevent
+    libgmp-dev # for gdb
+    gperf # for proxygen
+    git # for fbthrift
     libedit-dev libpcre3-dev automake bison # for swig
 )
 
@@ -26,6 +32,7 @@ TOOLCHAIN_RUN_DEPS=(
     libcurl4 # for cmake
     libreadline7 # for cmake and llvm
     libffi6 libxml2 # for llvm
+    libssl-dev # for libevent
 )
 
 MEMGRAPH_BUILD_DEPS=(
@@ -46,6 +53,7 @@ MEMGRAPH_BUILD_DEPS=(
     mono-runtime mono-mcs zip unzip default-jdk-headless # for driver tests
     dotnet-sdk-3.1 golang nodejs npm
     autoconf # for jemalloc code generation
+    libtool  # for protobuf code generation
 )
 
 list() {

environment/os/debian-11.sh

@@ -16,6 +16,12 @@ TOOLCHAIN_BUILD_DEPS=(
     libreadline-dev # for cmake and llvm
     libffi-dev libxml2-dev # for llvm
     libedit-dev libpcre3-dev automake bison # for swig
+    curl # snappy
+    file # for libunwind
+    libssl-dev # for libevent
+    libgmp-dev
+    gperf # for proxygen
+    git # for fbthrift
 )
 
 TOOLCHAIN_RUN_DEPS=(
@@ -27,6 +33,7 @@ TOOLCHAIN_RUN_DEPS=(
     file # for CPack
     libreadline8 # for cmake and llvm
     libffi7 libxml2 # for llvm
+    libssl-dev # for libevent
 )
 
 MEMGRAPH_BUILD_DEPS=(
@@ -47,6 +54,7 @@ MEMGRAPH_BUILD_DEPS=(
     mono-runtime mono-mcs zip unzip default-jdk-headless # for driver tests
     dotnet-sdk-3.1 golang nodejs npm
     autoconf # for jemalloc code generation
+    libtool  # for protobuf code generation
 )
 
 list() {

environment/os/debian-9.sh

@@ -1,62 +0,0 @@
-#!/bin/bash
-
-set -Eeuo pipefail
-
-DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" >/dev/null 2>&1 && pwd )"
-source "$DIR/../util.sh"
-
-TOOLCHAIN_BUILD_DEPS=(
-    coreutils gcc g++ build-essential make # generic build tools
-    wget # used for archive download
-    gnupg # used for archive signature verification
-    tar gzip bzip2 xz-utils unzip # used for archive unpacking
-    zlib1g-dev # zlib library used for all builds
-    libexpat1-dev libipt-dev libbabeltrace-dev libbabeltrace-ctf-dev liblzma-dev python3-dev texinfo # for gdb
-    libcurl4-openssl-dev # for cmake
-    libreadline-dev # for cmake and llvm
-    libffi-dev libxml2-dev # for llvm
-    libedit-dev libpcre3-dev automake bison # for swig
-)
-
-TOOLCHAIN_RUN_DEPS=(
-    make # generic build tools
-    tar gzip bzip2 xz-utils # used for archive unpacking
-    zlib1g # zlib library used for all builds
-    libexpat1 libipt1 libbabeltrace1 libbabeltrace-ctf1 liblzma5 python3 # for gdb
-    libcurl3 # for cmake
-    libreadline7 # for cmake and llvm
-    libffi6 libxml2 # for llvm
-)
-
-MEMGRAPH_BUILD_DEPS=(
-    git # source code control
-    make pkg-config # build system
-    curl wget # for downloading libs
-    uuid-dev default-jre-headless # required by antlr
-    libreadline-dev # for memgraph console
-    libpython3-dev python3-dev # for query modules
-    libssl-dev
-    libseccomp-dev
-    python3 python-virtualenv python3-pip # for qa, macro_benchmark and stress tests
-    python3-yaml # for the configuration generator
-    libcurl4-openssl-dev # mg-requests
-    sbcl # for custom Lisp C++ preprocessing
-    doxygen graphviz # source documentation generators
-    mono-runtime mono-mcs nodejs zip unzip default-jdk-headless # for driver tests
-    autoconf # for jemalloc code generation
-)
-
-list() {
-    echo "$1"
-}
-
-check() {
-    check_all_dpkg "$1"
-}
-
-install() {
-    install_all_apt "$1"
-}
-
-deps=$2"[*]"
-"$1" "${!deps}"

environment/os/ubuntu-18.04.sh

@@ -16,6 +16,12 @@ TOOLCHAIN_BUILD_DEPS=(
     libcurl4-openssl-dev # cmake
     libreadline-dev # cmake and llvm
     libffi-dev libxml2-dev # llvm
+    curl # snappy
+    file
+    git # for thrift
+    libgmp-dev # for gdb
+    gperf # for proxygen
+    libssl-dev
     libedit-dev libpcre3-dev automake bison # swig
 )
 
@@ -27,6 +33,7 @@ TOOLCHAIN_RUN_DEPS=(
     libcurl4 # for cmake
     libreadline7 # for cmake and llvm
     libffi6 libxml2 # for llvm
+    libssl-dev # for libevent
 )
 
 MEMGRAPH_BUILD_DEPS=(
@@ -45,6 +52,7 @@ MEMGRAPH_BUILD_DEPS=(
     doxygen graphviz # source documentation generators
     mono-runtime mono-mcs nodejs zip unzip default-jdk-headless # driver tests
     autoconf # for jemalloc code generation
+    libtool  # for protobuf code generation
 )
 
 list() {

environment/os/ubuntu-20.04.sh

@@ -15,6 +15,12 @@ TOOLCHAIN_BUILD_DEPS=(
     libcurl4-openssl-dev # for cmake
     libreadline-dev # for cmake and llvm
     libffi-dev libxml2-dev # for llvm
+    curl # snappy
+    file
+    git # for thrift
+    libgmp-dev # for gdb
+    gperf # for proxygen
+    libssl-dev
     libedit-dev libpcre3-dev automake bison # for swig
 )
 
@@ -26,6 +32,7 @@ TOOLCHAIN_RUN_DEPS=(
     libcurl4 # for cmake
     libreadline8 # for cmake and llvm
     libffi7 libxml2 # for llvm
+    libssl-dev # for libevent
 )
 
 MEMGRAPH_BUILD_DEPS=(
@@ -46,6 +53,7 @@ MEMGRAPH_BUILD_DEPS=(
     mono-runtime mono-mcs zip unzip default-jdk-headless # for driver tests
     dotnet-sdk-3.1 golang nodejs npm
     autoconf # for jemalloc code generation
+    libtool  # for protobuf code generation
 )
 
 list() {

environment/toolchain/proxygen.diff

@@ -0,0 +1,2 @@
+24d23
+< find_dependency(mvfst)

environment/toolchain/snappy.diff

@@ -0,0 +1,16 @@
+55,57c55,57
+<   # Disable RTTI.
+<   string(REGEX REPLACE "/GR" "" CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS}")
+<   set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /GR-")
+---
+>   # # Disable RTTI.
+>   # string(REGEX REPLACE "/GR" "" CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS}")
+>   # set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /GR-")
+80,82c80,82
+<   # Disable RTTI.
+<   string(REGEX REPLACE "-frtti" "" CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS}")
+<   set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fno-rtti")
+---
+>   # # Disable RTTI.
+>   # string(REGEX REPLACE "-frtti" "" CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS}")
+>   # set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fno-rtti")

environment/toolchain/xz_pgp.txt

@@ -0,0 +1,75 @@
+-----BEGIN PGP PUBLIC KEY BLOCK-----
+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+=asTC
+-----END PGP PUBLIC KEY BLOCK-----

include/mg_procedure.h

@@ -1405,38 +1405,76 @@ int mgp_must_abort(struct mgp_graph *graph);
 
 /// @}
 
-/// @name Kafka message API
-/// Currently the API below is for kafka only but in the future
-/// mgp_message and mgp_messages might be generic to support
-/// other streaming systems.
+/// @name Stream Source message API
+/// API for accessing specific data contained in a mgp_message
+/// used for defining transformation procedures.
+/// Not all methods are available for all stream sources
+/// so make sure that your transformation procedure can be used
+/// for a specific source, i.e. only valid methods are used.
 ///@{
 
-/// A single Kafka message
+/// A single Stream source message
 struct mgp_message;
 
-/// A list of Kafka messages
+/// A list of Stream source messages
 struct mgp_messages;
 
+/// Stream source type.
+enum mgp_source_type {
+  KAFKA,
+  PULSAR,
+};
+
+/// Get the type of the stream source that produced the message.
+enum mgp_error mgp_message_source_type(struct mgp_message *message, enum mgp_source_type *result);
+
 /// Payload is not null terminated and not a string but rather a byte array.
 /// You need to call mgp_message_payload_size() first, to read the size of
 /// the payload.
+/// Supported stream sources:
+///   - Kafka
+///   - Pulsar
+/// Return MGP_ERROR_INVALID_ARGUMENT if the message is from an unsupported stream source.
 enum mgp_error mgp_message_payload(struct mgp_message *message, const char **result);
 
 /// Get the payload size
+/// Supported stream sources:
+///   - Kafka
+///   - Pulsar
+/// Return MGP_ERROR_INVALID_ARGUMENT if the message is from an unsupported stream source.
 enum mgp_error mgp_message_payload_size(struct mgp_message *message, size_t *result);
 
 /// Get the name of topic
+/// Supported stream sources:
+///   - Kafka
+///   - Pulsar
+/// Return MGP_ERROR_INVALID_ARGUMENT if the message is from an unsupported stream source.
 enum mgp_error mgp_message_topic_name(struct mgp_message *message, const char **result);
 
 /// Get the key of mgp_message as a byte array
+/// Supported stream sources:
+///   - Kafka
+/// Return MGP_ERROR_INVALID_ARGUMENT if the message is from an unsupported stream source.
 enum mgp_error mgp_message_key(struct mgp_message *message, const char **result);
 
 /// Get the key size of mgp_message
+/// Supported stream sources:
+///   - Kafka
+/// Return MGP_ERROR_INVALID_ARGUMENT if the message is from an unsupported stream source.
 enum mgp_error mgp_message_key_size(struct mgp_message *message, size_t *result);
 
 /// Get the timestamp of mgp_message as a byte array
+/// Supported stream sources:
+///   - Kafka
+/// Return MGP_ERROR_INVALID_ARGUMENT if the message is from an unsupported stream source.
 enum mgp_error mgp_message_timestamp(struct mgp_message *message, int64_t *result);
 
+/// Get the message offset from a message.
+/// Supported stream sources:
+///   - Kafka
+/// Return MGP_ERROR_INVALID_ARGUMENT if the message is from an unsupported stream source.
+enum mgp_error mgp_message_offset(struct mgp_message *message, int64_t *result);
+
 /// Get the number of messages contained in the mgp_messages list
 /// Current implementation always returns without errors.
 enum mgp_error mgp_messages_size(struct mgp_messages *message, size_t *result);

include/mgp.py

@@ -1260,6 +1260,9 @@ class InvalidMessageError(Exception):
     pass
 
 
+SOURCE_TYPE_KAFKA = _mgp.SOURCE_TYPE_KAFKA
+SOURCE_TYPE_PULSAR = _mgp.SOURCE_TYPE_PULSAR
+
 class Message:
     """Represents a message from a stream."""
     __slots__ = ('_message',)
@@ -1280,26 +1283,73 @@ class Message:
         """Return True if `self` is in valid context and may be used."""
         return self._message.is_valid()
 
+    def source_type(self) -> str:
+        """
+        Supported in all stream sources
+
+        Raise InvalidArgumentError if the message is from an unsupported stream source.
+        """
+        if not self.is_valid():
+            raise InvalidMessageError()
+        return self._message.source_type()
+
     def payload(self) -> bytes:
+        """
+        Supported stream sources:
+          - Kafka
+          - Pulsar
+
+        Raise InvalidArgumentError if the message is from an unsupported stream source.
+        """
         if not self.is_valid():
             raise InvalidMessageError()
         return self._message.payload()
 
     def topic_name(self) -> str:
+        """
+        Supported stream sources:
+          - Kafka
+          - Pulsar
+
+        Raise InvalidArgumentError if the message is from an unsupported stream source.
+        """
         if not self.is_valid():
             raise InvalidMessageError()
         return self._message.topic_name()
 
     def key(self) -> bytes:
+        """
+        Supported stream sources:
+          - Kafka
+
+        Raise InvalidArgumentError if the message is from an unsupported stream source.
+        """
         if not self.is_valid():
             raise InvalidMessageError()
         return self._message.key()
 
     def timestamp(self) -> int:
+        """
+        Supported stream sources:
+          - Kafka
+
+        Raise InvalidArgumentError if the message is from an unsupported stream source.
+        """
         if not self.is_valid():
             raise InvalidMessageError()
         return self._message.timestamp()
 
+    def offset(self) -> int:
+        """
+        Supported stream sources:
+          - Kafka
+
+        Raise InvalidArgumentError if the message is from an unsupported stream source.
+        """
+        if not self.is_valid():
+            raise InvalidMessageError()
+        return self._message.offset()
+
 
 class InvalidMessagesError(Exception):
     """Signals using a messages instance outside of the registered transformation."""

libs/.gitignore

@@ -5,3 +5,4 @@
 !CMakeLists.txt
 !__main.cpp
 !jemalloc.cmake
+!pulsar.patch

libs/CMakeLists.txt

@@ -8,6 +8,15 @@ if (NPROC EQUAL 0)
   set(NPROC 1)
 endif()
 
+find_package(Boost 1.78 REQUIRED)
+find_package(BZip2 1.0.6 REQUIRED)
+find_package(Threads REQUIRED)
+set(GFLAGS_NOTHREADS OFF)
+find_package(gflags REQUIRED)
+find_package(fmt 8.0.1)
+find_package(Jemalloc REQUIRED)
+find_package(ZLIB 1.2.11 REQUIRED)
+
 set(LIB_DIR ${CMAKE_CURRENT_SOURCE_DIR})
 
 # convenience functions
@@ -48,7 +57,7 @@ endfunction(import_library)
 # INSTALL_COMMAND arguments.
 function(add_external_project name)
   set(options NO_C_COMPILER)
-  set(one_value_kwargs SOURCE_DIR)
+  set(one_value_kwargs SOURCE_DIR BUILD_IN_SOURCE)
   set(multi_value_kwargs CMAKE_ARGS DEPENDS INSTALL_COMMAND BUILD_COMMAND
       CONFIGURE_COMMAND)
   cmake_parse_arguments(KW "${options}" "${one_value_kwargs}" "${multi_value_kwargs}" ${ARGN})
@@ -56,11 +65,16 @@ function(add_external_project name)
   if (KW_SOURCE_DIR)
     set(source_dir ${KW_SOURCE_DIR})
   endif()
+  set(build_in_source 0)
+  if (KW_BUILD_IN_SOURCE)
+    set(build_in_source ${KW_BUILD_IN_SOURCE})
+  endif()
   if (NOT KW_NO_C_COMPILER)
     set(KW_CMAKE_ARGS -DCMAKE_C_COMPILER=${CMAKE_C_COMPILER} ${KW_CMAKE_ARGS})
   endif()
   ExternalProject_Add(${name}-proj DEPENDS ${KW_DEPENDS}
                       PREFIX ${source_dir} SOURCE_DIR ${source_dir}
+                      BUILD_IN_SOURCE ${build_in_source}
                       CONFIGURE_COMMAND ${KW_CONFIGURE_COMMAND}
                       CMAKE_ARGS -DCMAKE_BUILD_TYPE=Release
                                  -DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER}
@@ -97,17 +111,12 @@ import_external_library(antlr4 STATIC
 
 # Setup google benchmark.
 import_external_library(benchmark STATIC
-  ${CMAKE_CURRENT_SOURCE_DIR}/benchmark/lib/libbenchmark.a
+  ${CMAKE_CURRENT_SOURCE_DIR}/benchmark/${CMAKE_INSTALL_LIBDIR}/libbenchmark.a
   ${CMAKE_CURRENT_SOURCE_DIR}/benchmark/include
   # Skip testing. The tests don't compile with Clang 8.
   CMAKE_ARGS -DBENCHMARK_ENABLE_TESTING=OFF)
 
-# setup fmt format
-import_external_library(fmt STATIC
-  ${CMAKE_CURRENT_SOURCE_DIR}/fmt/${CMAKE_INSTALL_LIBDIR}/libfmt.a
-  ${CMAKE_CURRENT_SOURCE_DIR}/fmt/include
-  # Skip testing.
-  CMAKE_ARGS -DFMT_TEST=OFF)
+include(FetchContent)
 
 # setup rapidcheck (it cannot be external, since it doesn't have install
 # target)
@@ -134,55 +143,19 @@ import_library(gtest_main STATIC ${GTEST_MAIN_LIBRARY} ${GTEST_INCLUDE_DIR} gtes
 import_library(gmock STATIC ${GMOCK_LIBRARY} ${GTEST_INCLUDE_DIR} gtest-proj)
 import_library(gmock_main STATIC ${GMOCK_MAIN_LIBRARY} ${GTEST_INCLUDE_DIR} gtest-proj)
 
-# setup google flags
-set(GFLAGS_NO_FILENAMES "0")
-if ("${CMAKE_BUILD_TYPE}" MATCHES "^(R|r)(E|e)(L|l).+")
-  set(GFLAGS_NO_FILENAMES "1")
-endif()
-
-# setup google flags
-import_external_library(gflags STATIC
-  ${CMAKE_CURRENT_SOURCE_DIR}/gflags/lib/libgflags.a
-  ${CMAKE_CURRENT_SOURCE_DIR}/gflags/include
-  # Not needed, since gflags is C++ only.
-  NO_C_COMPILER
-  # Don't register installation in ~/.cmake
-  CMAKE_ARGS -DREGISTER_INSTALL_PREFIX=OFF
-             -DBUILD_gflags_nothreads_LIB=OFF
-             -DGFLAGS_NO_FILENAMES=${GFLAGS_NO_FILENAMES})
-
 # Setup cppitertools
 import_header_library(cppitertools ${CMAKE_CURRENT_SOURCE_DIR})
 
 # Setup json
 import_header_library(json ${CMAKE_CURRENT_SOURCE_DIR})
 
-# Setup bzip2
-import_external_library(bzip2 STATIC
-    ${CMAKE_CURRENT_SOURCE_DIR}/bzip2/libbz2.a
-    ${CMAKE_CURRENT_SOURCE_DIR}/bzip2
-    # bzip2's Makefile has -g CFLAG which is redundant
-    CONFIGURE_COMMAND sed -i "s/-Wall -Winline -O2 -g/-Wall -Winline -O2/g" ${CMAKE_CURRENT_SOURCE_DIR}/bzip2/Makefile
-    BUILD_COMMAND make -C ${CMAKE_CURRENT_SOURCE_DIR}/bzip2
-                       CC=${CMAKE_C_COMPILER}
-                       CXX=${CMAKE_CXX_COMPILER}
-    INSTALL_COMMAND true)
-
-# Setup zlib
-import_external_library(zlib STATIC
-    ${CMAKE_CURRENT_SOURCE_DIR}/zlib/lib/libz.a
-    ${CMAKE_CURRENT_SOURCE_DIR}/zlib
-    CMAKE_ARGS -DCMAKE_SKIP_INSTALL_ALL_DEPENDENCY=true
-    BUILD_COMMAND $(MAKE) zlibstatic)
-
 # Setup RocksDB
 import_external_library(rocksdb STATIC
   ${CMAKE_CURRENT_SOURCE_DIR}/rocksdb/lib/librocksdb.a
   ${CMAKE_CURRENT_SOURCE_DIR}/rocksdb/include
-  DEPENDS gflags-proj
-  CMAKE_ARGS -Dgflags_DIR=${CMAKE_CURRENT_SOURCE_DIR}/gflags/lib/cmake/gflags
-             -DUSE_RTTI=ON
+  CMAKE_ARGS -DUSE_RTTI=ON
              -DWITH_TESTS=OFF
+             -DGFLAGS_NOTHREADS=OFF
              -DCMAKE_INSTALL_LIBDIR=lib
              -DCMAKE_SKIP_INSTALL_ALL_DEPENDENCY=true
   BUILD_COMMAND $(MAKE) rocksdb)
@@ -217,12 +190,11 @@ add_external_project(mgconsole
 add_custom_target(mgconsole DEPENDS mgconsole-proj)
 
 # Setup spdlog
-import_external_library(spdlog STATIC
-  ${CMAKE_CURRENT_SOURCE_DIR}/spdlog/${CMAKE_INSTALL_LIBDIR}/libspdlog.a
-  ${CMAKE_CURRENT_SOURCE_DIR}/spdlog/include
-  BUILD_COMMAND $(MAKE) spdlog)
+set(SPDLOG_FMT_EXTERNAL ON)
+FetchContent_Declare(spdlog
+  SOURCE_DIR ${CMAKE_CURRENT_SOURCE_DIR}/spdlog)
 
-include(jemalloc.cmake)
+FetchContent_MakeAvailable(spdlog)
 
 # Setup librdkafka.
 import_external_library(librdkafka STATIC
@@ -231,14 +203,45 @@ import_external_library(librdkafka STATIC
   CMAKE_ARGS -DRDKAFKA_BUILD_STATIC=ON
              -DRDKAFKA_BUILD_EXAMPLES=OFF
              -DRDKAFKA_BUILD_TESTS=OFF
+             -DWITH_ZSTD=OFF
+             -DENABLE_LZ4_EXT=OFF
              -DCMAKE_INSTALL_LIBDIR=lib
              -DWITH_SSL=ON
              # If we want SASL, we need to install it on build machines
              -DWITH_SASL=OFF)
-target_link_libraries(librdkafka INTERFACE ${OPENSSL_LIBRARIES} zlib)
+target_link_libraries(librdkafka INTERFACE ${OPENSSL_LIBRARIES} ZLIB::ZLIB)
 
 import_library(librdkafka++ STATIC
   ${CMAKE_CURRENT_SOURCE_DIR}/librdkafka/lib/librdkafka++.a
   ${CMAKE_CURRENT_SOURCE_DIR}/librdkafka/include
 )
 target_link_libraries(librdkafka++ INTERFACE librdkafka)
+
+set(PROTOBUF_ROOT ${CMAKE_CURRENT_SOURCE_DIR}/protobuf/lib)
+import_external_library(protobuf STATIC
+  ${PROTOBUF_ROOT}/lib/libprotobuf.a
+  ${PROTOBUF_ROOT}/include
+  BUILD_IN_SOURCE 1
+  CONFIGURE_COMMAND true)
+
+import_external_library(pulsar STATIC
+  ${CMAKE_CURRENT_SOURCE_DIR}/pulsar/pulsar-client-cpp/lib/libpulsarwithdeps.a
+  ${CMAKE_CURRENT_SOURCE_DIR}/pulsar/install/include
+  BUILD_IN_SOURCE 1
+  CONFIGURE_COMMAND cmake pulsar-client-cpp
+                            -DCMAKE_INSTALL_PREFIX=${CMAKE_CURRENT_SOURCE_DIR}/pulsar/install
+                            -DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER}
+                            -DCMAKE_C_COMPILER=${CMAKE_C_COMPILER}
+                            -DBUILD_DYNAMIC_LIB=OFF
+                            -DBUILD_STATIC_LIB=ON
+                            -DBUILD_TESTS=OFF
+                            -DLINK_STATIC=ON
+                            -DPROTOC_PATH=${PROTOBUF_ROOT}/bin/protoc
+                            -DBOOST_ROOT=${BOOST_ROOT}
+                            -DCMAKE_PREFIX_PATH=${PROTOBUF_ROOT}
+                            -DProtobuf_INCLUDE_DIRS=${PROTOBUF_ROOT}/include
+                            -DBUILD_PYTHON_WRAPPER=OFF
+                            -DBUILD_PERF_TOOLS=OFF
+                            -DUSE_LOG4CXX=OFF
+ BUILD_COMMAND $(MAKE) pulsarStaticWithDeps)
+add_dependencies(pulsar-proj protobuf)

libs/antlr4.patch

@@ -0,0 +1,43 @@
+diff --git a/runtime/Cpp/runtime/CMakeLists.txt b/runtime/Cpp/runtime/CMakeLists.txt
+index a8503bb..11362cf 100644
+--- a/runtime/Cpp/runtime/CMakeLists.txt
++++ b/runtime/Cpp/runtime/CMakeLists.txt
+@@ -5,8 +5,8 @@ set(THIRDPARTY_DIR ${CMAKE_BINARY_DIR}/runtime/thirdparty)
+ set(UTFCPP_DIR ${THIRDPARTY_DIR}/utfcpp)
+ ExternalProject_Add(
+   utfcpp
+-  GIT_REPOSITORY        "git://github.com/nemtrif/utfcpp"
+-  GIT_TAG               "v3.1.1"
++  GIT_REPOSITORY        "https://github.com/nemtrif/utfcpp"
++  GIT_TAG               "v3.2.1"
+   SOURCE_DIR            ${UTFCPP_DIR}
+   UPDATE_DISCONNECTED   1
+   CMAKE_ARGS            -DCMAKE_INSTALL_PREFIX=${UTFCPP_DIR}/install -Dgtest_force_shared_crt=ON
+@@ -118,7 +118,7 @@ set_target_properties(antlr4_static
+                                  ARCHIVE_OUTPUT_DIRECTORY ${LIB_OUTPUT_DIR}
+                                  COMPILE_FLAGS "${disabled_compile_warnings} ${extra_static_compile_flags}")
+ 
+-install(TARGETS antlr4_shared
++install(TARGETS antlr4_shared OPTIONAL
+         DESTINATION lib
+         EXPORT antlr4-targets)
+ install(TARGETS antlr4_static
+diff --git a/runtime/Cpp/runtime/src/support/Any.h b/runtime/Cpp/runtime/src/support/Any.h
+index 468db98..65a473b 100644
+--- a/runtime/Cpp/runtime/src/support/Any.h
++++ b/runtime/Cpp/runtime/src/support/Any.h
+@@ -122,12 +122,12 @@ private:
+     }
+ 
+   private:
+-    template<int N = 0, typename std::enable_if<N == N && std::is_nothrow_copy_constructible<T>::value, int>::type = 0>
++    template<int N = 0, typename std::enable_if<N == N && std::is_copy_constructible<T>::value, int>::type = 0>
+     Base* clone() const {
+       return new Derived<T>(value);
+     }
+ 
+-    template<int N = 0, typename std::enable_if<N == N && !std::is_nothrow_copy_constructible<T>::value, int>::type = 0>
++    template<int N = 0, typename std::enable_if<N == N && !std::is_copy_constructible<T>::value, int>::type = 0>
+     Base* clone() const {
+       return nullptr;
+     }

libs/rocksdb.patch

@@ -0,0 +1,21 @@
+diff --git a/CMakeLists.txt b/CMakeLists.txt
+index 6761929..6a369af 100644
+--- a/CMakeLists.txt
++++ b/CMakeLists.txt
+@@ -220,6 +220,7 @@ else()
+       set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -momit-leaf-frame-pointer")
+     endif()
+   endif()
++  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wno-deprecated-copy -Wno-unused-but-set-variable")
+ endif()
+ 
+ include(CheckCCompilerFlag)
+@@ -997,7 +998,7 @@ if(NOT WIN32 OR ROCKSDB_INSTALL_ON_WINDOWS)
+ 
+   if(ROCKSDB_BUILD_SHARED)
+     install(
+-      TARGETS ${ROCKSDB_SHARED_LIB}
++      TARGETS ${ROCKSDB_SHARED_LIB} OPTIONAL
+       EXPORT RocksDBTargets
+       COMPONENT runtime
+       ARCHIVE DESTINATION "${CMAKE_INSTALL_LIBDIR}"

libs/setup.sh

@@ -14,39 +14,49 @@ clone () {
   local git_repo=$1
   local dir_name=$2
   local checkout_id=$3
-  shift 3
+  local shallow=$4
+  shift 4
   # Clone if there's no repo.
   if [[ ! -d "$dir_name" ]]; then
     echo "Cloning from $git_repo"
     # If the clone fails, it doesn't make sense to continue with the function
     # execution but the whole script should continue executing because we might
     # clone the same repo from a different source.
-    git clone "$git_repo" "$dir_name" || return 1
+
+    if [ "$shallow" = true ]; then
+      git clone --depth 1 --branch "$checkout_id" "$git_repo" "$dir_name" || return 1
+    else
+      git clone "$git_repo" "$dir_name" || return 1
+    fi
   fi
   pushd "$dir_name"
-  # Just fetch new commits from remote repository. Don't merge/pull them in, so
-  # that we don't clobber local modifications.
-  git fetch
   # Check whether we have any local changes which need to be preserved.
   local local_changes=true
   if git diff --no-ext-diff --quiet && git diff --no-ext-diff --cached --quiet; then
     local_changes=false
   fi
-  # Stash regardless of local_changes, so that a user gets a message on stdout.
-  git stash
-  # Checkout the primary commit (there's no need to pull/merge).
-  # The checkout fail should exit this script immediately because the target
-  # commit is not there and that will most likely create build-time errors.
-  git checkout "$checkout_id" || exit 1
-  # Apply any optional cherry pick fixes.
-  while [[ $# -ne 0 ]]; do
-    local cherry_pick_id=$1
-    shift
-    # The cherry-pick fail should exit this script immediately because the
-    # target commit is not there and that will most likely create build-time
-    # errors.
-    git cherry-pick -n "$cherry_pick_id" || exit 1
-  done
+
+  if [ "$shallow" = false ]; then
+    # Stash regardless of local_changes, so that a user gets a message on stdout.
+    git stash
+    # Just fetch new commits from remote repository. Don't merge/pull them in, so
+    # that we don't clobber local modifications.
+    git fetch
+    # Checkout the primary commit (there's no need to pull/merge).
+    # The checkout fail should exit this script immediately because the target
+    # commit is not there and that will most likely create build-time errors.
+    git checkout "$checkout_id" || exit 1
+    # Apply any optional cherry pick fixes.
+    while [[ $# -ne 0 ]]; do
+      local cherry_pick_id=$1
+      shift
+      # The cherry-pick fail should exit this script immediately because the
+      # target commit is not there and that will most likely create build-time
+      # errors.
+      git cherry-pick -n "$cherry_pick_id" || exit 1
+    done
+  fi
+
   # Reapply any local changes.
   if [[ $local_changes == true ]]; then
     git stash pop
@@ -70,12 +80,13 @@ repo_clone_try_double () {
     secondary_url="$2"
     folder_name="$3"
     ref="$4"
+    shallow="${5:-false}"
     echo "Cloning primary from $primary_url secondary from $secondary_url"
     if [ -z "$primary_url" ]; then echo "Primary should not be empty." && exit 1; fi
     if [ -z "$secondary_url" ]; then echo "Secondary should not be empty." && exit 1; fi
     if [ -z "$folder_name" ]; then echo "Clone folder should not be empty." && exit 1; fi
     if [ -z "$ref" ]; then echo "Git clone ref should not be empty." && exit 1; fi
-    clone "$primary_url" "$folder_name" "$ref" || clone "$secondary_url" "$folder_name" "$ref" || exit 1
+    clone "$primary_url" "$folder_name" "$ref" "$shallow" || clone "$secondary_url" "$folder_name" "$ref" "$shallow" || exit 1
     echo ""
 }
 
@@ -113,6 +124,9 @@ declare -A primary_urls=(
   ["nlohmann"]="http://$local_cache_host/file/nlohmann/json/b3e5cb7f20dcc5c806e418df34324eca60d17d4e/single_include/nlohmann/json.hpp"
   ["neo4j"]="http://$local_cache_host/file/neo4j-community-3.2.3-unix.tar.gz"
   ["librdkafka"]="http://$local_cache_host/git/librdkafka.git"
+  ["protobuf"]="http://$local_cache_host/git/protobuf.git"
+  ["boost"]="http://$local_cache_host/file/boost_1_77_0.tar.gz"
+  ["pulsar"]="http://$local_cache_host/git/pulsar.git"
 )
 
 # The goal of secondary urls is to have links to the "source of truth" of
@@ -140,45 +154,35 @@ declare -A secondary_urls=(
   ["nlohmann"]="https://raw.githubusercontent.com/nlohmann/json/b3e5cb7f20dcc5c806e418df34324eca60d17d4e/single_include/nlohmann/json.hpp"
   ["neo4j"]="https://s3-eu-west-1.amazonaws.com/deps.memgraph.io/neo4j-community-3.2.3-unix.tar.gz"
   ["librdkafka"]="https://github.com/edenhill/librdkafka.git"
+  ["protobuf"]="https://github.com/protocolbuffers/protobuf.git"
+  ["boost"]="https://boostorg.jfrog.io/artifactory/main/release/1.77.0/source/boost_1_77_0.tar.gz"
+  ["pulsar"]="https://github.com/apache/pulsar.git"
 )
 
 # antlr
 file_get_try_double "${primary_urls[antlr4-generator]}" "${secondary_urls[antlr4-generator]}"
 
-antlr4_tag="5e5b6d35b4183fd330102c40947b95c4b5c6abb5" # v4.9.2
-repo_clone_try_double "${primary_urls[antlr4-code]}" "${secondary_urls[antlr4-code]}" "antlr4" "$antlr4_tag"
-# remove shared library from install dependencies
-sed -i 's/install(TARGETS antlr4_shared/install(TARGETS antlr4_shared OPTIONAL/' antlr4/runtime/Cpp/runtime/CMakeLists.txt
-# fix issue https://github.com/antlr/antlr4/issues/3194 - should update Antlr commit once the PR related to the issue gets merged
-sed -i 's/std::is_nothrow_copy_constructible/std::is_copy_constructible/' antlr4/runtime/Cpp/runtime/src/support/Any.h
-# replace the utf8cpp version which is used because the older one uses gtest that doesn't 
-# compile with the newer compilers because of uninitialized variable
-# the newer utf8cpp switched to ftest
-sed -i 's/v3.1.1/v3.2.1/' antlr4/runtime/Cpp/runtime/CMakeLists.txt
+antlr4_tag="4.9.2" # v4.9.2
+repo_clone_try_double "${primary_urls[antlr4-code]}" "${secondary_urls[antlr4-code]}" "antlr4" "$antlr4_tag" true
+pushd antlr4
+git apply ../antlr4.patch
+popd
 
 # cppitertools v2.0 2019-12-23
 cppitertools_ref="cb3635456bdb531121b82b4d2e3afc7ae1f56d47"
 repo_clone_try_double "${primary_urls[cppitertools]}" "${secondary_urls[cppitertools]}" "cppitertools" "$cppitertools_ref"
 
-# fmt
-fmt_tag="7bdf0628b1276379886c7f6dda2cef2b3b374f0b" # (2020-11-25)
-repo_clone_try_double "${primary_urls[fmt]}" "${secondary_urls[fmt]}" "fmt" "$fmt_tag"
-
 # rapidcheck
 rapidcheck_tag="7bc7d302191a4f3d0bf005692677126136e02f60" # (2020-05-04)
 repo_clone_try_double "${primary_urls[rapidcheck]}" "${secondary_urls[rapidcheck]}" "rapidcheck" "$rapidcheck_tag"
 
 # google benchmark
-benchmark_tag="4f8bfeae470950ef005327973f15b0044eceaceb" # v1.1.0
-repo_clone_try_double "${primary_urls[gbenchmark]}" "${secondary_urls[gbenchmark]}" "benchmark" "$benchmark_tag"
+benchmark_tag="v1.6.0"
+repo_clone_try_double "${primary_urls[gbenchmark]}" "${secondary_urls[gbenchmark]}" "benchmark" "$benchmark_tag" true
 
 # google test
-googletest_tag="ec44c6c1675c25b9827aacd08c02433cccde7780" # v1.8.0
-repo_clone_try_double "${primary_urls[gtest]}" "${secondary_urls[gtest]}" "googletest" "$googletest_tag"
-
-# google flags
-gflags_tag="b37ceb03a0e56c9f15ce80409438a555f8a67b7c" # custom version (May 6, 2017)
-repo_clone_try_double "${primary_urls[gflags]}" "${secondary_urls[gflags]}" "gflags" "$gflags_tag"
+googletest_tag="release-1.8.0"
+repo_clone_try_double "${primary_urls[gtest]}" "${secondary_urls[gtest]}" "googletest" "$googletest_tag" true
 
 # libbcrypt
 libbcrypt_tag="8aa32ad94ebe06b76853b0767c910c9fbf7ccef4" # custom version (Dec 16, 2016)
@@ -198,22 +202,15 @@ cd json
 file_get_try_double "${primary_urls[nlohmann]}" "${secondary_urls[nlohmann]}"
 cd ..
 
-bzip2_tag="0405487e2b1de738e7f1c8afb50d19cf44e8d580"  # v1.0.6 (May 26, 2011)
-repo_clone_try_double "${primary_urls[bzip2]}" "${secondary_urls[bzip2]}" "bzip2" "$bzip2_tag"
-
-zlib_tag="cacf7f1d4e3d44d871b605da3b647f07d718623f" # v1.2.11.
-repo_clone_try_double "${primary_urls[zlib]}" "${secondary_urls[zlib]}" "zlib" "$zlib_tag"
-# remove shared library from install dependencies
-sed -i 's/install(TARGETS zlib zlibstatic/install(TARGETS zlibstatic/g' zlib/CMakeLists.txt
-
-rocksdb_tag="f3e33549c151f30ac4eb7c22356c6d0331f37652" # (2020-10-14)
-repo_clone_try_double "${primary_urls[rocksdb]}" "${secondary_urls[rocksdb]}" "rocksdb" "$rocksdb_tag"
-# remove shared library from install dependencies
-sed -i 's/TARGETS ${ROCKSDB_SHARED_LIB}/TARGETS ${ROCKSDB_SHARED_LIB} OPTIONAL/' rocksdb/CMakeLists.txt
+rocksdb_tag="v6.14.6" # (2020-10-14)
+repo_clone_try_double "${primary_urls[rocksdb]}" "${secondary_urls[rocksdb]}" "rocksdb" "$rocksdb_tag" true
+pushd rocksdb
+git apply ../rocksdb.patch
+popd
 
 # mgclient
 mgclient_tag="v1.3.0" # (2021-09-23)
-repo_clone_try_double "${primary_urls[mgclient]}" "${secondary_urls[mgclient]}" "mgclient" "$mgclient_tag"
+repo_clone_try_double "${primary_urls[mgclient]}" "${secondary_urls[mgclient]}" "mgclient" "$mgclient_tag" true
 sed -i 's/\${CMAKE_INSTALL_LIBDIR}/lib/' mgclient/src/CMakeLists.txt
 
 # pymgclient
@@ -222,30 +219,25 @@ repo_clone_try_double "${primary_urls[pymgclient]}" "${secondary_urls[pymgclient
 
 # mgconsole
 mgconsole_tag="v1.1.0" # (2021-10-07)
-repo_clone_try_double "${primary_urls[mgconsole]}" "${secondary_urls[mgconsole]}" "mgconsole" "$mgconsole_tag"
+repo_clone_try_double "${primary_urls[mgconsole]}" "${secondary_urls[mgconsole]}" "mgconsole" "$mgconsole_tag" true
 
-spdlog_tag="46d418164dd4cd9822cf8ca62a116a3f71569241" # (2020-12-01)
-repo_clone_try_double "${primary_urls[spdlog]}" "${secondary_urls[spdlog]}" "spdlog" "$spdlog_tag"
-
-jemalloc_tag="ea6b3e973b477b8061e0076bb257dbd7f3faa756" # (2021-02-11)
-repo_clone_try_double "${primary_urls[jemalloc]}" "${secondary_urls[jemalloc]}" "jemalloc" "$jemalloc_tag"
-pushd jemalloc
-# ThreadPool select job randomly, and there can be some threads that had been
-# performed some memory heavy task before and will be inactive for some time,
-# but until it will became active again, the memory will not be freed since by
-# default each thread has it's own arena, but there should be not more then
-# 4*CPU arenas (see opt.nareans description).
-#
-# By enabling percpu_arena number of arenas limited to number of CPUs and hence
-# this problem should go away.
-#
-# muzzy_decay_ms -- use MADV_FREE when available on newer Linuxes, to
-# avoid spurious latencies and additional work associated with
-# MADV_DONTNEED. See
-# https://github.com/ClickHouse/ClickHouse/issues/11121 for motivation.
-./autogen.sh --with-malloc-conf="percpu_arena:percpu,oversize_threshold:0,muzzy_decay_ms:5000,dirty_decay_ms:5000"
-popd
+spdlog_tag="v1.9.2" # (2021-08-12)
+repo_clone_try_double "${primary_urls[spdlog]}" "${secondary_urls[spdlog]}" "spdlog" "$spdlog_tag" true
 
 # librdkafka
 librdkafka_tag="v1.7.0" # (2021-05-06)
-repo_clone_try_double "${primary_urls[librdkafka]}" "${secondary_urls[librdkafka]}" "librdkafka" "$librdkafka_tag"
+repo_clone_try_double "${primary_urls[librdkafka]}" "${secondary_urls[librdkafka]}" "librdkafka" "$librdkafka_tag" true
+
+# protobuf
+protobuf_tag="v3.12.4"
+repo_clone_try_double "${primary_urls[protobuf]}" "${secondary_urls[protobuf]}" "protobuf" "$protobuf_tag" true
+pushd protobuf
+./autogen.sh && ./configure CC=clang CXX=clang++ --prefix=$(pwd)/lib
+popd
+
+#pulsar
+pulsar_tag="v2.8.1"
+repo_clone_try_double "${primary_urls[pulsar]}" "${secondary_urls[pulsar]}" "pulsar" "$pulsar_tag" true
+pushd pulsar
+git apply ../pulsar.patch
+popd

licenses/BSL.txt

@@ -36,7 +36,7 @@ ADDITIONAL USE GRANT:   You may use the Licensed Work in accordance with the
                         3. using the Licensed Work to create a work or solution
                         which competes (or might reasonably be expected to
                         compete) with the Licensed Work.
-CHANGE DATE:            2025-10-12
+CHANGE DATE:            2025-12-08
 CHANGE LICENSE:         Apache License, Version 2.0
 
 For information about alternative licensing arrangements, please visit: https://memgraph.com/legal.

release/README.md

@@ -1,4 +0,0 @@
-# Memgraph Release
-
-Please continue in
-[Notion](https://www.notion.so/memgraph/Release-621733518c82456ebc119d2a025b5dd3).

release/docker/memgraph.dockerfile

@@ -1,14 +1,14 @@
-FROM debian:buster
+FROM debian:bullseye
 # NOTE: If you change the base distro update release/package as well.
 
 ARG deb_release
 
 RUN apt-get update && apt-get install -y \
-    openssl libcurl4 libssl1.1 libseccomp2 python3 libpython3.7 python3-pip \
+    openssl libcurl4 libssl1.1 libseccomp2 python3 libpython3.9 python3-pip \
     --no-install-recommends \
   && rm -rf /var/lib/apt/lists/* /tmp/* /var/tmp/*
 
-RUN pip3 install networkx==2.4 numpy==1.19.2 scipy==1.5.2
+RUN pip3 install networkx==2.4 numpy==1.21.4 scipy==1.7.3
 
 COPY ${deb_release} /
 

release/package/run.sh

@@ -5,7 +5,7 @@ set -Eeuo pipefail
 SCRIPT_DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )"
 SUPPORTED_OS=(centos-7 centos-8 debian-9 debian-10 debian-11 ubuntu-18.04 ubuntu-20.04)
 PROJECT_ROOT="$SCRIPT_DIR/../.."
-TOOLCHAIN_VERSION="toolchain-v3"
+TOOLCHAIN_VERSION="toolchain-v4"
 ACTIVATE_TOOLCHAIN="source /opt/${TOOLCHAIN_VERSION}/activate"
 HOST_OUTPUT_DIR="$PROJECT_ROOT/build/output"
 
@@ -72,7 +72,7 @@ make_package () {
     docker exec "$build_container" bash -c "/memgraph/environment/os/$os.sh install MEMGRAPH_BUILD_DEPS"
 
     echo "Building targeted package..."
-    docker exec "$build_container" bash -c "cd /memgraph && ./init"
+    docker exec "$build_container" bash -c "cd /memgraph && $ACTIVATE_TOOLCHAIN && ./init"
     docker exec "$build_container" bash -c "cd $container_build_dir && rm -rf ./*"
     docker exec "$build_container" bash -c "cd $container_build_dir && $ACTIVATE_TOOLCHAIN && cmake -DCMAKE_BUILD_TYPE=release $telemetry_id_override_flag .."
     # ' is used instead of " because we need to run make within the allowed
@@ -100,8 +100,8 @@ case "$1" in
     ;;
 
     docker)
-        # NOTE: Docker is build on top of Debian 10 package.
-        based_on_os="debian-10"
+        # NOTE: Docker is build on top of Debian 11 package.
+        based_on_os="debian-11"
         # shellcheck disable=SC2012
         last_package_name=$(cd "$HOST_OUTPUT_DIR/$based_on_os" && ls -t memgraph* | head -1)
         docker_build_folder="$PROJECT_ROOT/release/docker"

src/CMakeLists.txt

@@ -37,7 +37,7 @@ set(mg_single_node_v2_sources
 )
 
 set(mg_single_node_v2_libs stdc++fs Threads::Threads
-    telemetry_lib mg-query mg-communication mg-memory mg-utils mg-auth)
+    telemetry_lib mg-query mg-communication mg-memory mg-utils mg-auth mg-license mg-settings)
 if (MG_ENTERPRISE)
   # These are enterprise subsystems
   set(mg_single_node_v2_libs ${mg_single_node_v2_libs} mg-audit)

src/audit/CMakeLists.txt

@@ -1,5 +1,8 @@
 set(audit_src_files log.cpp)
 
+find_package(fmt REQUIRED)
+find_package(gflags REQUIRED)
+
 add_library(mg-audit STATIC ${audit_src_files})
-target_link_libraries(mg-audit json gflags fmt)
+target_link_libraries(mg-audit json gflags fmt::fmt)
 target_link_libraries(mg-audit mg-utils mg-storage-v2)

src/auth/CMakeLists.txt

@@ -5,9 +5,12 @@ set(auth_src_files
     module.cpp)
 
 find_package(Seccomp REQUIRED)
+find_package(fmt REQUIRED)
+find_package(gflags REQUIRED)
+
 
 add_library(mg-auth STATIC ${auth_src_files})
-target_link_libraries(mg-auth json libbcrypt gflags fmt)
+target_link_libraries(mg-auth json libbcrypt gflags fmt::fmt)
 target_link_libraries(mg-auth mg-utils mg-kvstore)
 
 target_link_libraries(mg-auth ${Seccomp_LIBRARIES})

src/communication/CMakeLists.txt

@@ -1,3 +1,6 @@
+find_package(fmt REQUIRED)
+find_package(gflags REQUIRED)
+
 set(communication_src_files
     bolt/v1/value.cpp
     buffer.cpp
@@ -7,7 +10,7 @@ set(communication_src_files
     init.cpp)
 
 add_library(mg-communication STATIC ${communication_src_files})
-target_link_libraries(mg-communication Threads::Threads mg-utils mg-io fmt gflags)
+target_link_libraries(mg-communication Threads::Threads mg-utils mg-io fmt::fmt gflags)
 
 find_package(OpenSSL REQUIRED)
 target_link_libraries(mg-communication ${OPENSSL_LIBRARIES})

src/communication/bolt/v1/codes.hpp

@@ -31,6 +31,7 @@ enum class Signature : uint8_t {
   Begin = 0x11,
   Commit = 0x12,
   Rollback = 0x13,
+  Route = 0x66,
 
   Record = 0x71,
   Success = 0x70,

src/communication/bolt/v1/constants.hpp

@@ -28,7 +28,7 @@ static constexpr size_t kChunkWholeSize = kChunkHeaderSize + kChunkMaxDataSize;
  */
 static constexpr size_t kHandshakeSize = 20;
 
-static constexpr uint16_t kSupportedVersions[3] = {0x0100, 0x0400, 0x0401};
+static constexpr uint16_t kSupportedVersions[] = {0x0100, 0x0400, 0x0401, 0x0403};
 
 static constexpr int kPullAll = -1;
 static constexpr int kPullLast = -1;

src/communication/bolt/v1/states/executing.hpp

@@ -12,6 +12,7 @@
 #pragma once
 
 #include <map>
+#include <memory>
 #include <new>
 #include <string>
 
@@ -19,339 +20,69 @@
 #include "communication/bolt/v1/constants.hpp"
 #include "communication/bolt/v1/exceptions.hpp"
 #include "communication/bolt/v1/state.hpp"
+#include "communication/bolt/v1/states/handlers.hpp"
 #include "communication/bolt/v1/value.hpp"
-#include "communication/exceptions.hpp"
 #include "utils/likely.hpp"
 #include "utils/logging.hpp"
 #include "utils/message.hpp"
 
 namespace communication::bolt {
 
-// TODO (mferencevic): revise these error messages
-inline std::pair<std::string, std::string> ExceptionToErrorMessage(const std::exception &e) {
-  if (auto *verbose = dynamic_cast<const VerboseError *>(&e)) {
-    return {verbose->code(), verbose->what()};
-  }
-  if (dynamic_cast<const ClientError *>(&e)) {
-    // Clients expect 4 strings separated by dots. First being database name
-    // (for example: Neo, Memgraph...), second being either ClientError,
-    // TransientError or DatabaseError (or ClientNotification for warnings).
-    // ClientError means wrong query, do not retry. DatabaseError means
-    // something wrong in database, do not retry. TransientError means query
-    // failed, but if retried it may succeed, retry it.
-    //
-    // Third and fourth strings being namespace and specific error name.
-    // It is not really important what we put there since we don't expect
-    // any special handling of specific exceptions on client side, but we
-    // need to make sure that we don't accidentally return some exception
-    // name which clients handle in a special way. For example, if client
-    // receives *.TransientError.Transaction.Terminate it will not rerun
-    // query even though TransientError was returned, because of Neo's
-    // semantics of that error.
-    return {"Memgraph.ClientError.MemgraphError.MemgraphError", e.what()};
-  }
-  if (dynamic_cast<const utils::BasicException *>(&e)) {
-    // Exception not derived from QueryException was thrown which means that
-    // database probably aborted transaction because of some timeout,
-    // deadlock, serialization error or something similar. We return
-    // TransientError since retry of same transaction could succeed.
-    return {"Memgraph.TransientError.MemgraphError.MemgraphError", e.what()};
-  }
-  if (dynamic_cast<const std::bad_alloc *>(&e)) {
-    // std::bad_alloc was thrown, God knows in which state is database ->
-    // terminate.
-    LOG_FATAL("Memgraph is out of memory");
-  }
-  // All exceptions used in memgraph are derived from BasicException. Since
-  // we caught some other exception we don't know what is going on. Return
-  // DatabaseError, log real message and return generic string.
-  spdlog::error(
-      utils::MessageWithLink("Unknown exception occurred during query execution {}.", e.what(), "https://memgr.ph/unknown"));
-  return {"Memgraph.DatabaseError.MemgraphError.MemgraphError",
-          "An unknown exception occurred, this is unexpected. Real message "
-          "should be in database logs."};
-}
-
 template <typename TSession>
-inline State HandleFailure(TSession &session, const std::exception &e) {
-  spdlog::trace("Error message: {}", e.what());
-  if (const auto *p = dynamic_cast<const utils::StacktraceException *>(&e)) {
-    spdlog::trace("Error trace: {}", p->trace());
-  }
-  session.encoder_buffer_.Clear();
-  auto code_message = ExceptionToErrorMessage(e);
-  bool fail_sent = session.encoder_.MessageFailure({{"code", code_message.first}, {"message", code_message.second}});
-  if (!fail_sent) {
-    spdlog::trace("Couldn't send failure message!");
-    return State::Close;
-  }
-  return State::Error;
-}
-
-template <typename TSession>
-State HandleRun(TSession &session, State state, Marker marker) {
-  const std::map<std::string, Value> kEmptyFields = {{"fields", std::vector<Value>{}}};
-
-  const auto expected_marker = session.version_.major == 1 ? Marker::TinyStruct2 : Marker::TinyStruct3;
-  if (marker != expected_marker) {
-    spdlog::trace("Expected {} marker, but received 0x{:02X}!",
-                  session.version_.major == 1 ? "TinyStruct2" : "TinyStruct3", utils::UnderlyingCast(marker));
-    return State::Close;
-  }
-
-  Value query, params, extra;
-  if (!session.decoder_.ReadValue(&query, Value::Type::String)) {
-    spdlog::trace("Couldn't read query string!");
-    return State::Close;
-  }
-
-  if (!session.decoder_.ReadValue(&params, Value::Type::Map)) {
-    spdlog::trace("Couldn't read parameters!");
-    return State::Close;
-  }
-
-  if (session.version_.major == 4) {
-    if (!session.decoder_.ReadValue(&extra, Value::Type::Map)) {
-      spdlog::trace("Couldn't read extra field!");
-    }
-  }
-
-  if (state != State::Idle) {
-    // Client could potentially recover if we move to error state, but there is
-    // no legitimate situation in which well working client would end up in this
-    // situation.
-    spdlog::trace("Unexpected RUN command!");
-    return State::Close;
-  }
-
-  DMG_ASSERT(!session.encoder_buffer_.HasData(), "There should be no data to write in this state");
-
-  spdlog::debug("[Run] '{}'", query.ValueString());
-
-  try {
-    // Interpret can throw.
-    auto [header, qid] = session.Interpret(query.ValueString(), params.ValueMap());
-    // Convert std::string to Value
-    std::vector<Value> vec;
-    std::map<std::string, Value> data;
-    vec.reserve(header.size());
-    for (auto &i : header) vec.emplace_back(std::move(i));
-    data.emplace("fields", std::move(vec));
-    // Send the header.
-    if (!session.encoder_.MessageSuccess(data)) {
-      spdlog::trace("Couldn't send query header!");
+State RunHandlerV1(Signature signature, TSession &session, State state, Marker marker) {
+  switch (signature) {
+    case Signature::Run:
+      return HandleRunV1<TSession>(session, state, marker);
+    case Signature::Pull:
+      return HandlePullV1<TSession>(session, state, marker);
+    case Signature::Discard:
+      return HandleDiscardV1<TSession>(session, state, marker);
+    case Signature::Reset:
+      return HandleReset<TSession>(session, marker);
+    default:
+      spdlog::trace("Unrecognized signature received (0x{:02X})!", utils::UnderlyingCast(signature));
       return State::Close;
-    }
-    return State::Result;
-  } catch (const std::exception &e) {
-    return HandleFailure(session, e);
   }
 }
 
-namespace detail {
-template <bool is_pull, typename TSession>
-State HandlePullDiscard(TSession &session, State state, Marker marker) {
-  const auto expected_marker = session.version_.major == 1 ? Marker::TinyStruct : Marker::TinyStruct1;
-  if (marker != expected_marker) {
-    spdlog::trace("Expected {} marker, but received 0x{:02X}!",
-                  session.version_.major == 1 ? "TinyStruct" : "TinyStruct1", utils::UnderlyingCast(marker));
-    return State::Close;
-  }
-
-  if (state != State::Result) {
-    if constexpr (is_pull) {
-      spdlog::trace("Unexpected PULL!");
-    } else {
-      spdlog::trace("Unexpected DISCARD!");
-    }
-    // Same as `unexpected RUN` case.
-    return State::Close;
-  }
-
-  try {
-    std::optional<int> n;
-    std::optional<int> qid;
-
-    if (session.version_.major == 4) {
-      Value extra;
-      if (!session.decoder_.ReadValue(&extra, Value::Type::Map)) {
-        spdlog::trace("Couldn't read extra field!");
-      }
-      const auto &extra_map = extra.ValueMap();
-      if (extra_map.count("n")) {
-        if (const auto n_value = extra_map.at("n").ValueInt(); n_value != kPullAll) {
-          n = n_value;
-        }
-      }
-
-      if (extra_map.count("qid")) {
-        if (const auto qid_value = extra_map.at("qid").ValueInt(); qid_value != kPullLast) {
-          qid = qid_value;
-        }
+template <typename TSession, int bolt_minor = 0>
+State RunHandlerV4(Signature signature, TSession &session, State state, Marker marker) {
+  switch (signature) {
+    case Signature::Run:
+      return HandleRunV4<TSession>(session, state, marker);
+    case Signature::Pull:
+      return HandlePullV4<TSession>(session, state, marker);
+    case Signature::Discard:
+      return HandleDiscardV4<TSession>(session, state, marker);
+    case Signature::Reset:
+      return HandleReset<TSession>(session, marker);
+    case Signature::Begin:
+      return HandleBegin<TSession>(session, state, marker);
+    case Signature::Commit:
+      return HandleCommit<TSession>(session, state, marker);
+    case Signature::Goodbye:
+      return HandleGoodbye<TSession>();
+    case Signature::Rollback:
+      return HandleRollback<TSession>(session, state, marker);
+    case Signature::Noop: {
+      if constexpr (bolt_minor >= 1) {
+        return HandleNoop<TSession>(state);
+      } else {
+        spdlog::trace("Supported only in bolt v4.1");
+        return State::Close;
       }
     }
-
-    std::map<std::string, Value> summary;
-    if constexpr (is_pull) {
-      // Pull can throw.
-      summary = session.Pull(&session.encoder_, n, qid);
-    } else {
-      summary = session.Discard(n, qid);
+    case Signature::Route: {
+      if constexpr (bolt_minor >= 3) {
+        if (signature == Signature::Route) return HandleRoute<TSession>(session);
+      } else {
+        spdlog::trace("Supported only in bolt v4.3");
+        return State::Close;
+      }
     }
-
-    if (!session.encoder_.MessageSuccess(summary)) {
-      spdlog::trace("Couldn't send query summary!");
+    default:
+      spdlog::trace("Unrecognized signature received (0x{:02X})!", utils::UnderlyingCast(signature));
       return State::Close;
-    }
-
-    if (summary.count("has_more") && summary.at("has_more").ValueBool()) {
-      return State::Result;
-    }
-
-    return State::Idle;
-  } catch (const std::exception &e) {
-    return HandleFailure(session, e);
-  }
-}
-}  // namespace detail
-
-template <typename Session>
-State HandlePull(Session &session, State state, Marker marker) {
-  return detail::HandlePullDiscard<true>(session, state, marker);
-}
-
-template <typename Session>
-State HandleDiscard(Session &session, State state, Marker marker) {
-  return detail::HandlePullDiscard<false>(session, state, marker);
-}
-
-template <typename Session>
-State HandleReset(Session &session, State, Marker marker) {
-  // IMPORTANT: This implementation of the Bolt RESET command isn't fully
-  // compliant to the protocol definition. In the protocol it is defined
-  // that this command should immediately stop any running commands and
-  // reset the session to a clean state. That means that we should always
-  // make a look-ahead for the RESET command before processing anything.
-  // Our implementation, for now, does everything in a blocking fashion
-  // so we cannot simply "kill" a transaction while it is running. So
-  // now this command only resets the session to a clean state. It
-  // does not IGNORE running and pending commands as it should.
-  if (marker != Marker::TinyStruct) {
-    spdlog::trace("Expected TinyStruct marker, but received 0x{:02X}!", utils::UnderlyingCast(marker));
-    return State::Close;
-  }
-
-  // Clear all pending data and send a success message.
-  session.encoder_buffer_.Clear();
-
-  if (!session.encoder_.MessageSuccess()) {
-    spdlog::trace("Couldn't send success message!");
-    return State::Close;
-  }
-
-  session.Abort();
-
-  return State::Idle;
-}
-
-template <typename Session>
-State HandleBegin(Session &session, State state, Marker marker) {
-  if (session.version_.major == 1) {
-    spdlog::trace("BEGIN messsage not supported in Bolt v1!");
-    return State::Close;
-  }
-
-  if (marker != Marker::TinyStruct1) {
-    spdlog::trace("Expected TinyStruct1 marker, but received 0x{:02x}!", utils::UnderlyingCast(marker));
-    return State::Close;
-  }
-
-  Value extra;
-  if (!session.decoder_.ReadValue(&extra, Value::Type::Map)) {
-    spdlog::trace("Couldn't read extra fields!");
-    return State::Close;
-  }
-
-  if (state != State::Idle) {
-    spdlog::trace("Unexpected BEGIN command!");
-    return State::Close;
-  }
-
-  DMG_ASSERT(!session.encoder_buffer_.HasData(), "There should be no data to write in this state");
-
-  if (!session.encoder_.MessageSuccess({})) {
-    spdlog::trace("Couldn't send success message!");
-    return State::Close;
-  }
-
-  try {
-    session.BeginTransaction();
-  } catch (const std::exception &e) {
-    return HandleFailure(session, e);
-  }
-
-  return State::Idle;
-}
-
-template <typename Session>
-State HandleCommit(Session &session, State state, Marker marker) {
-  if (session.version_.major == 1) {
-    spdlog::trace("COMMIT messsage not supported in Bolt v1!");
-    return State::Close;
-  }
-
-  if (marker != Marker::TinyStruct) {
-    spdlog::trace("Expected TinyStruct marker, but received 0x{:02x}!", utils::UnderlyingCast(marker));
-    return State::Close;
-  }
-
-  if (state != State::Idle) {
-    spdlog::trace("Unexpected COMMIT command!");
-    return State::Close;
-  }
-
-  DMG_ASSERT(!session.encoder_buffer_.HasData(), "There should be no data to write in this state");
-
-  try {
-    if (!session.encoder_.MessageSuccess({})) {
-      spdlog::trace("Couldn't send success message!");
-      return State::Close;
-    }
-    session.CommitTransaction();
-    return State::Idle;
-  } catch (const std::exception &e) {
-    return HandleFailure(session, e);
-  }
-}
-
-template <typename Session>
-State HandleRollback(Session &session, State state, Marker marker) {
-  if (session.version_.major == 1) {
-    spdlog::trace("ROLLBACK messsage not supported in Bolt v1!");
-    return State::Close;
-  }
-
-  if (marker != Marker::TinyStruct) {
-    spdlog::trace("Expected TinyStruct marker, but received 0x{:02x}!", utils::UnderlyingCast(marker));
-    return State::Close;
-  }
-
-  if (state != State::Idle) {
-    spdlog::trace("Unexpected ROLLBACK command!");
-    return State::Close;
-  }
-
-  DMG_ASSERT(!session.encoder_buffer_.HasData(), "There should be no data to write in this state");
-
-  try {
-    if (!session.encoder_.MessageSuccess({})) {
-      spdlog::trace("Couldn't send success message!");
-      return State::Close;
-    }
-    session.RollbackTransaction();
-    return State::Idle;
-  } catch (const std::exception &e) {
-    return HandleFailure(session, e);
   }
 }
 
@@ -361,8 +92,8 @@ State HandleRollback(Session &session, State state, Marker marker) {
  * It executes: RUN, PULL_ALL, DISCARD_ALL & RESET.
  * @param session the session that should be used for the run
  */
-template <typename Session>
-State StateExecutingRun(Session &session, State state) {
+template <typename TSession>
+State StateExecutingRun(TSession &session, State state) {
   Marker marker;
   Signature signature;
   if (!session.decoder_.ReadMessageHeader(&signature, &marker)) {
@@ -370,30 +101,21 @@ State StateExecutingRun(Session &session, State state) {
     return State::Close;
   }
 
-  if (UNLIKELY(signature == Signature::Noop && session.version_.major == 4 && session.version_.minor == 1)) {
-    spdlog::trace("Received NOOP message");
-    return state;
-  }
-
-  if (signature == Signature::Run) {
-    return HandleRun(session, state, marker);
-  } else if (signature == Signature::Pull) {
-    return HandlePull(session, state, marker);
-  } else if (signature == Signature::Discard) {
-    return HandleDiscard(session, state, marker);
-  } else if (signature == Signature::Begin) {
-    return HandleBegin(session, state, marker);
-  } else if (signature == Signature::Commit) {
-    return HandleCommit(session, state, marker);
-  } else if (signature == Signature::Rollback) {
-    return HandleRollback(session, state, marker);
-  } else if (signature == Signature::Reset) {
-    return HandleReset(session, state, marker);
-  } else if (signature == Signature::Goodbye && session.version_.major != 1) {
-    throw SessionClosedException("Closing connection.");
-  } else {
-    spdlog::trace("Unrecognized signature received (0x{:02X})!", utils::UnderlyingCast(signature));
-    return State::Close;
+  switch (session.version_.major) {
+    case 1:
+      return RunHandlerV1(signature, session, state, marker);
+    case 4: {
+      if (session.version_.minor >= 3) {
+        return RunHandlerV4<TSession, 3>(signature, session, state, marker);
+      }
+      if (session.version_.minor >= 1) {
+        return RunHandlerV4<TSession, 1>(signature, session, state, marker);
+      }
+      return RunHandlerV4<TSession>(signature, session, state, marker);
+    }
+    default:
+      spdlog::trace("Unsupported bolt version:{}.{})!", session.version_.major, session.version_.minor);
+      return State::Close;
   }
 }
 }  // namespace communication::bolt

src/communication/bolt/v1/states/handlers.hpp

@@ -0,0 +1,418 @@
+// 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.
+
+#pragma once
+
+#include <map>
+#include <string>
+#include <string_view>
+#include <vector>
+
+#include "communication/bolt/v1/codes.hpp"
+#include "communication/bolt/v1/constants.hpp"
+#include "communication/bolt/v1/state.hpp"
+#include "communication/bolt/v1/value.hpp"
+#include "communication/exceptions.hpp"
+#include "utils/logging.hpp"
+#include "utils/message.hpp"
+
+namespace communication::bolt {
+// TODO: Revise these error messages
+inline std::pair<std::string, std::string> ExceptionToErrorMessage(const std::exception &e) {
+  if (const auto *verbose = dynamic_cast<const VerboseError *>(&e)) {
+    return {verbose->code(), verbose->what()};
+  }
+  if (dynamic_cast<const ClientError *>(&e)) {
+    // Clients expect 4 strings separated by dots. First being database name
+    // (for example: Neo, Memgraph...), second being either ClientError,
+    // TransientError or DatabaseError (or ClientNotification for warnings).
+    // ClientError means wrong query, do not retry. DatabaseError means
+    // something wrong in database, do not retry. TransientError means query
+    // failed, but if retried it may succeed, retry it.
+    //
+    // Third and fourth strings being namespace and specific error name.
+    // It is not really important what we put there since we don't expect
+    // any special handling of specific exceptions on client side, but we
+    // need to make sure that we don't accidentally return some exception
+    // name which clients handle in a special way. For example, if client
+    // receives *.TransientError.Transaction.Terminate it will not rerun
+    // query even though TransientError was returned, because of Neo's
+    // semantics of that error.
+    return {"Memgraph.ClientError.MemgraphError.MemgraphError", e.what()};
+  }
+  if (dynamic_cast<const utils::BasicException *>(&e)) {
+    // Exception not derived from QueryException was thrown which means that
+    // database probably aborted transaction because of some timeout,
+    // deadlock, serialization error or something similar. We return
+    // TransientError since retry of same transaction could succeed.
+    return {"Memgraph.TransientError.MemgraphError.MemgraphError", e.what()};
+  }
+  if (dynamic_cast<const std::bad_alloc *>(&e)) {
+    // std::bad_alloc was thrown, God knows in which state is database ->
+    // terminate.
+    LOG_FATAL("Memgraph is out of memory");
+  }
+  // All exceptions used in memgraph are derived from BasicException. Since
+  // we caught some other exception we don't know what is going on. Return
+  // DatabaseError, log real message and return generic string.
+  spdlog::error(utils::MessageWithLink("Unknown exception occurred during query execution {}.", e.what(),
+                                       "https://memgr.ph/unknown"));
+  return {"Memgraph.DatabaseError.MemgraphError.MemgraphError",
+          "An unknown exception occurred, this is unexpected. Real message "
+          "should be in database logs."};
+}
+
+namespace details {
+
+template <typename TSession>
+State HandleRun(TSession &session, const State state, const Value &query, const Value &params) {
+  if (state != State::Idle) {
+    // Client could potentially recover if we move to error state, but there is
+    // no legitimate situation in which well working client would end up in this
+    // situation.
+    spdlog::trace("Unexpected RUN command!");
+    return State::Close;
+  }
+
+  DMG_ASSERT(!session.encoder_buffer_.HasData(), "There should be no data to write in this state");
+
+  spdlog::debug("[Run] '{}'", query.ValueString());
+
+  try {
+    // Interpret can throw.
+    const auto [header, qid] = session.Interpret(query.ValueString(), params.ValueMap());
+    // Convert std::string to Value
+    std::vector<Value> vec;
+    std::map<std::string, Value> data;
+    vec.reserve(header.size());
+    for (auto &i : header) vec.emplace_back(std::move(i));
+    data.emplace("fields", std::move(vec));
+    // Send the header.
+    if (!session.encoder_.MessageSuccess(data)) {
+      spdlog::trace("Couldn't send query header!");
+      return State::Close;
+    }
+    return State::Result;
+  } catch (const std::exception &e) {
+    return HandleFailure(session, e);
+  }
+}
+
+template <bool is_pull, typename TSession>
+State HandlePullDiscard(TSession &session, std::optional<int> n, std::optional<int> qid) {
+  try {
+    std::map<std::string, Value> summary;
+    if constexpr (is_pull) {
+      // Pull can throw.
+      summary = session.Pull(&session.encoder_, n, qid);
+    } else {
+      summary = session.Discard(n, qid);
+    }
+
+    if (!session.encoder_.MessageSuccess(summary)) {
+      spdlog::trace("Couldn't send query summary!");
+      return State::Close;
+    }
+
+    if (summary.count("has_more") && summary.at("has_more").ValueBool()) {
+      return State::Result;
+    }
+
+    return State::Idle;
+  } catch (const std::exception &e) {
+    return HandleFailure(session, e);
+  }
+}
+
+template <bool is_pull, typename TSession>
+State HandlePullDiscardV1(TSession &session, const State state, const Marker marker) {
+  const auto expected_marker = Marker::TinyStruct;
+  if (marker != expected_marker) {
+    spdlog::trace("Expected {} marker, but received 0x{:02X}!", "TinyStruct", utils::UnderlyingCast(marker));
+    return State::Close;
+  }
+
+  if (state != State::Result) {
+    if constexpr (is_pull) {
+      spdlog::trace("Unexpected PULL!");
+    } else {
+      spdlog::trace("Unexpected DISCARD!");
+    }
+    // Same as `unexpected RUN` case.
+    return State::Close;
+  }
+
+  return HandlePullDiscard<is_pull, TSession>(session, std::nullopt, std::nullopt);
+}
+
+template <bool is_pull, typename TSession>
+State HandlePullDiscardV4(TSession &session, const State state, const Marker marker) {
+  const auto expected_marker = Marker::TinyStruct1;
+  if (marker != expected_marker) {
+    spdlog::trace("Expected {} marker, but received 0x{:02X}!", "TinyStruct1", utils::UnderlyingCast(marker));
+    return State::Close;
+  }
+
+  if (state != State::Result) {
+    if constexpr (is_pull) {
+      spdlog::trace("Unexpected PULL!");
+    } else {
+      spdlog::trace("Unexpected DISCARD!");
+    }
+    // Same as `unexpected RUN` case.
+    return State::Close;
+  }
+  std::optional<int> n;
+  std::optional<int> qid;
+  Value extra;
+  if (!session.decoder_.ReadValue(&extra, Value::Type::Map)) {
+    spdlog::trace("Couldn't read extra field!");
+  }
+  const auto &extra_map = extra.ValueMap();
+  if (extra_map.count("n")) {
+    if (const auto n_value = extra_map.at("n").ValueInt(); n_value != kPullAll) {
+      n = n_value;
+    }
+  }
+
+  if (extra_map.count("qid")) {
+    if (const auto qid_value = extra_map.at("qid").ValueInt(); qid_value != kPullLast) {
+      qid = qid_value;
+    }
+  }
+  return HandlePullDiscard<is_pull, TSession>(session, n, qid);
+}
+}  // namespace details
+
+template <typename TSession>
+inline State HandleFailure(TSession &session, const std::exception &e) {
+  spdlog::trace("Error message: {}", e.what());
+  if (const auto *p = dynamic_cast<const utils::StacktraceException *>(&e)) {
+    spdlog::trace("Error trace: {}", p->trace());
+  }
+  session.encoder_buffer_.Clear();
+  auto code_message = ExceptionToErrorMessage(e);
+  bool fail_sent = session.encoder_.MessageFailure({{"code", code_message.first}, {"message", code_message.second}});
+  if (!fail_sent) {
+    spdlog::trace("Couldn't send failure message!");
+    return State::Close;
+  }
+  return State::Error;
+}
+
+template <typename TSession>
+State HandleRunV1(TSession &session, const State state, const Marker marker) {
+  const auto expected_marker = Marker::TinyStruct2;
+  if (marker != expected_marker) {
+    spdlog::trace("Expected {} marker, but received 0x{:02X}!",
+                  session.version_.major == 1 ? "TinyStruct2" : "TinyStruct3", utils::UnderlyingCast(marker));
+    return State::Close;
+  }
+  Value query, params;
+  if (!session.decoder_.ReadValue(&query, Value::Type::String)) {
+    spdlog::trace("Couldn't read query string!");
+    return State::Close;
+  }
+
+  if (!session.decoder_.ReadValue(&params, Value::Type::Map)) {
+    spdlog::trace("Couldn't read parameters!");
+    return State::Close;
+  }
+
+  return details::HandleRun(session, state, query, params);
+}
+
+template <typename TSession>
+State HandleRunV4(TSession &session, const State state, const Marker marker) {
+  const auto expected_marker = Marker::TinyStruct3;
+  if (marker != expected_marker) {
+    spdlog::trace("Expected {} marker, but received 0x{:02X}!", "TinyStruct3", utils::UnderlyingCast(marker));
+    return State::Close;
+  }
+  Value query, params, extra;
+  if (!session.decoder_.ReadValue(&query, Value::Type::String)) {
+    spdlog::trace("Couldn't read query string!");
+    return State::Close;
+  }
+
+  if (!session.decoder_.ReadValue(&params, Value::Type::Map)) {
+    spdlog::trace("Couldn't read parameters!");
+    return State::Close;
+  }
+
+  // Even though this part seems unnecessary it is needed to move the buffer
+  if (!session.decoder_.ReadValue(&extra, Value::Type::Map)) {
+    spdlog::trace("Couldn't read extra field!");
+  }
+
+  return details::HandleRun(session, state, query, params);
+}
+
+template <typename TSession>
+State HandlePullV1(TSession &session, const State state, const Marker marker) {
+  return details::HandlePullDiscardV1<true>(session, state, marker);
+}
+
+template <typename TSession>
+State HandlePullV4(TSession &session, const State state, const Marker marker) {
+  return details::HandlePullDiscardV4<true>(session, state, marker);
+}
+
+template <typename TSession>
+State HandleDiscardV1(TSession &session, const State state, const Marker marker) {
+  return details::HandlePullDiscardV1<false>(session, state, marker);
+}
+
+template <typename TSession>
+State HandleDiscardV4(TSession &session, const State state, const Marker marker) {
+  return details::HandlePullDiscardV4<false>(session, state, marker);
+}
+
+template <typename TSession>
+State HandleReset(TSession &session, const Marker marker) {
+  // IMPORTANT: This implementation of the Bolt RESET command isn't fully
+  // compliant to the protocol definition. In the protocol it is defined
+  // that this command should immediately stop any running commands and
+  // reset the session to a clean state. That means that we should always
+  // make a look-ahead for the RESET command before processing anything.
+  // Our implementation, for now, does everything in a blocking fashion
+  // so we cannot simply "kill" a transaction while it is running. So
+  // now this command only resets the session to a clean state. It
+  // does not IGNORE running and pending commands as it should.
+  if (marker != Marker::TinyStruct) {
+    spdlog::trace("Expected TinyStruct marker, but received 0x{:02X}!", utils::UnderlyingCast(marker));
+    return State::Close;
+  }
+
+  // Clear all pending data and send a success message.
+  session.encoder_buffer_.Clear();
+
+  if (!session.encoder_.MessageSuccess()) {
+    spdlog::trace("Couldn't send success message!");
+    return State::Close;
+  }
+
+  session.Abort();
+
+  return State::Idle;
+}
+
+template <typename TSession>
+State HandleBegin(TSession &session, const State state, const Marker marker) {
+  if (marker != Marker::TinyStruct1) {
+    spdlog::trace("Expected TinyStruct1 marker, but received 0x{:02x}!", utils::UnderlyingCast(marker));
+    return State::Close;
+  }
+
+  Value extra;
+  if (!session.decoder_.ReadValue(&extra, Value::Type::Map)) {
+    spdlog::trace("Couldn't read extra fields!");
+    return State::Close;
+  }
+
+  if (state != State::Idle) {
+    spdlog::trace("Unexpected BEGIN command!");
+    return State::Close;
+  }
+
+  DMG_ASSERT(!session.encoder_buffer_.HasData(), "There should be no data to write in this state");
+
+  if (!session.encoder_.MessageSuccess({})) {
+    spdlog::trace("Couldn't send success message!");
+    return State::Close;
+  }
+
+  try {
+    session.BeginTransaction();
+  } catch (const std::exception &e) {
+    return HandleFailure(session, e);
+  }
+
+  return State::Idle;
+}
+
+template <typename TSession>
+State HandleCommit(TSession &session, const State state, const Marker marker) {
+  if (marker != Marker::TinyStruct) {
+    spdlog::trace("Expected TinyStruct marker, but received 0x{:02x}!", utils::UnderlyingCast(marker));
+    return State::Close;
+  }
+
+  if (state != State::Idle) {
+    spdlog::trace("Unexpected COMMIT command!");
+    return State::Close;
+  }
+
+  DMG_ASSERT(!session.encoder_buffer_.HasData(), "There should be no data to write in this state");
+
+  try {
+    if (!session.encoder_.MessageSuccess({})) {
+      spdlog::trace("Couldn't send success message!");
+      return State::Close;
+    }
+    session.CommitTransaction();
+    return State::Idle;
+  } catch (const std::exception &e) {
+    return HandleFailure(session, e);
+  }
+}
+
+template <typename TSession>
+State HandleRollback(TSession &session, const State state, const Marker marker) {
+  if (marker != Marker::TinyStruct) {
+    spdlog::trace("Expected TinyStruct marker, but received 0x{:02x}!", utils::UnderlyingCast(marker));
+    return State::Close;
+  }
+
+  if (state != State::Idle) {
+    spdlog::trace("Unexpected ROLLBACK command!");
+    return State::Close;
+  }
+
+  DMG_ASSERT(!session.encoder_buffer_.HasData(), "There should be no data to write in this state");
+
+  try {
+    if (!session.encoder_.MessageSuccess({})) {
+      spdlog::trace("Couldn't send success message!");
+      return State::Close;
+    }
+    session.RollbackTransaction();
+    return State::Idle;
+  } catch (const std::exception &e) {
+    return HandleFailure(session, e);
+  }
+}
+
+template <typename TSession>
+State HandleNoop(const State state) {
+  spdlog::trace("Received NOOP message");
+  return state;
+}
+
+template <typename TSession>
+State HandleGoodbye() {
+  throw SessionClosedException("Closing connection.");
+}
+
+template <typename TSession>
+State HandleRoute(TSession &session) {
+  // Route message is not implemented since it is neo4j specific, therefore we
+  // will receive it an inform user that there is no implementation.
+  session.encoder_buffer_.Clear();
+  bool fail_sent =
+      session.encoder_.MessageFailure({{"code", 66}, {"message", "Route message not supported in Memgraph!"}});
+  if (!fail_sent) {
+    spdlog::trace("Couldn't send failure message!");
+    return State::Close;
+  }
+  return State::Error;
+}
+}  // namespace communication::bolt

src/communication/bolt/v1/states/handshake.hpp

@@ -12,6 +12,9 @@
 #pragma once
 
 #include <fmt/format.h>
+#include <algorithm>
+#include <cstdint>
+#include <iterator>
 
 #include "communication/bolt/v1/codes.hpp"
 #include "communication/bolt/v1/constants.hpp"
@@ -21,6 +24,35 @@
 
 namespace communication::bolt {
 
+inline bool CopyProtocolInformationIfSupported(uint16_t version, uint8_t *protocol) {
+  const auto *supported_version = std::find(std::begin(kSupportedVersions), std::end(kSupportedVersions), version);
+  if (supported_version != std::end(kSupportedVersions)) {
+    std::memcpy(protocol, &version, sizeof(version));
+    return true;
+  }
+  return false;
+}
+
+inline bool CopyProtocolInformationIfSupportedWithOffset(auto data_position, uint8_t *protocol) {
+  struct bolt_range_version {
+    uint8_t offset;
+    uint8_t minor;
+    uint8_t major;
+  } bolt_range_version;
+  std::memcpy(&bolt_range_version, data_position, sizeof(bolt_range_version));
+  if (bolt_range_version.major == 0 || bolt_range_version.minor == 0) return false;
+  bolt_range_version.offset = std::min(bolt_range_version.offset, bolt_range_version.minor);
+
+  for (uint8_t i{0U}; i <= bolt_range_version.offset; i++) {
+    uint8_t current_minor = bolt_range_version.minor - i;
+    if (CopyProtocolInformationIfSupported(static_cast<uint16_t>((bolt_range_version.major << 8U) + current_minor),
+                                           protocol)) {
+      return true;
+    }
+  }
+  return false;
+}
+
 /**
  * Handshake state run function
  * This function runs everything to make a Bolt handshake with the client.
@@ -29,7 +61,7 @@ namespace communication::bolt {
 template <typename TSession>
 State StateHandshakeRun(TSession &session) {
   auto precmp = std::memcmp(session.input_stream_.data(), kPreamble, sizeof(kPreamble));
-  if (UNLIKELY(precmp != 0)) {
+  if (precmp != 0) [[unlikely]] {
     spdlog::trace("Received a wrong preamble!");
     return State::Close;
   }
@@ -37,22 +69,22 @@ State StateHandshakeRun(TSession &session) {
   DMG_ASSERT(session.input_stream_.size() >= kHandshakeSize, "Wrong size of the handshake data!");
 
   auto dataPosition = session.input_stream_.data() + sizeof(kPreamble);
-
   uint8_t protocol[4] = {0x00};
-  for (int i = 0; i < 4 && !protocol[3]; ++i) {
-    dataPosition += 2;  // version is defined only by the last 2 bytes
 
-    uint16_t version = 0;
+  for (int i = 0; i < 4 && !protocol[3]; ++i) {
+    // If there is an offset defined (e.g. 0x00 0x03 0x03 0x04) the second byte
+    // That would enable the client to pick between 4.0 and 4.3 versions
+    // as per changes in handshake bolt protocol in v4.3
+    if (CopyProtocolInformationIfSupportedWithOffset(dataPosition + 1, protocol + 2)) break;
+
+    dataPosition += 2;  // version is defined only by the last 2 bytes
+    uint16_t version{0};
     std::memcpy(&version, dataPosition, sizeof(version));
     if (!version) {
       break;
     }
-
-    for (const auto supportedVersion : kSupportedVersions) {
-      if (supportedVersion == version) {
-        std::memcpy(protocol + 2, &version, sizeof(version));
-        break;
-      }
+    if (CopyProtocolInformationIfSupported(version, protocol + 2)) {
+      break;
     }
 
     dataPosition += 2;

src/communication/bolt/v1/states/init.hpp

@@ -12,6 +12,7 @@
 #pragma once
 
 #include <fmt/format.h>
+#include <optional>
 
 #include "communication/bolt/v1/codes.hpp"
 #include "communication/bolt/v1/state.hpp"
@@ -22,10 +23,45 @@
 
 namespace communication::bolt {
 
-namespace detail {
+namespace details {
 template <typename TSession>
-std::optional<Value> StateInitRunV1(TSession &session, const Marker marker) {
-  if (UNLIKELY(marker != Marker::TinyStruct2)) {
+std::optional<State> AuthenticateUser(TSession &session, Value &metadata) {
+  // Get authentication data.
+  auto &data = metadata.ValueMap();
+  if (!data.count("scheme")) {
+    spdlog::warn("The client didn't supply authentication information!");
+    return State::Close;
+  }
+  std::string username;
+  std::string password;
+  if (data["scheme"].ValueString() == "basic") {
+    if (!data.count("principal") || !data.count("credentials")) {
+      spdlog::warn("The client didn't supply authentication information!");
+      return State::Close;
+    }
+    username = data["principal"].ValueString();
+    password = data["credentials"].ValueString();
+  } else if (data["scheme"].ValueString() != "none") {
+    spdlog::warn("Unsupported authentication scheme: {}", data["scheme"].ValueString());
+    return State::Close;
+  }
+
+  // Authenticate the user.
+  if (!session.Authenticate(username, password)) {
+    if (!session.encoder_.MessageFailure(
+            {{"code", "Memgraph.ClientError.Security.Unauthenticated"}, {"message", "Authentication failure"}})) {
+      spdlog::trace("Couldn't send failure message to the client!");
+    }
+    // Throw an exception to indicate to the network stack that the session
+    // should be closed and cleaned up.
+    throw SessionClosedException("The client is not authenticated!");
+  }
+  return std::nullopt;
+}
+
+template <typename TSession>
+std::optional<Value> GetMetadataV1(TSession &session, const Marker marker) {
+  if (marker != Marker::TinyStruct2) [[unlikely]] {
     spdlog::trace("Expected TinyStruct2 marker, but received 0x{:02X}!", utils::UnderlyingCast(marker));
     spdlog::trace(
         "The client sent malformed data, but we are continuing "
@@ -53,8 +89,8 @@ std::optional<Value> StateInitRunV1(TSession &session, const Marker marker) {
 }
 
 template <typename TSession>
-std::optional<Value> StateInitRunV4(TSession &session, const Marker marker) {
-  if (UNLIKELY(marker != Marker::TinyStruct1)) {
+std::optional<Value> GetMetadataV4(TSession &session, const Marker marker) {
+  if (marker != Marker::TinyStruct1) [[unlikely]] {
     spdlog::trace("Expected TinyStruct1 marker, but received 0x{:02X}!", utils::UnderlyingCast(marker));
     spdlog::trace(
         "The client sent malformed data, but we are continuing "
@@ -80,15 +116,79 @@ std::optional<Value> StateInitRunV4(TSession &session, const Marker marker) {
 
   return metadata;
 }
-}  // namespace detail
+
+template <typename TSession>
+State SendSuccessMessage(TSession &session) {
+  // Neo4j's Java driver 4.1.1+ requires connection_id.
+  // The only usage in the mentioned version is for logging purposes.
+  // Because it's not critical for the regular usage of the driver
+  // we send a hardcoded value for now.
+  std::map<std::string, Value> metadata{{"connection_id", "bolt-1"}};
+  if (auto server_name = session.GetServerNameForInit(); server_name) {
+    metadata.insert({"server", *server_name});
+  }
+  bool success_sent = session.encoder_.MessageSuccess(metadata);
+  if (!success_sent) {
+    spdlog::trace("Couldn't send success message to the client!");
+    return State::Close;
+  }
+
+  return State::Idle;
+}
+
+template <typename TSession>
+State StateInitRunV1(TSession &session, const Marker marker, const Signature signature) {
+  if (signature != Signature::Init) [[unlikely]] {
+    spdlog::trace("Expected Init signature, but received 0x{:02X}!", utils::UnderlyingCast(signature));
+    return State::Close;
+  }
+
+  auto maybeMetadata = GetMetadataV1(session, marker);
+
+  if (!maybeMetadata) {
+    return State::Close;
+  }
+  if (auto result = AuthenticateUser(session, *maybeMetadata)) {
+    return result.value();
+  }
+
+  return SendSuccessMessage(session);
+}
+
+template <typename TSession, int bolt_minor = 0>
+State StateInitRunV4(TSession &session, Marker marker, Signature signature) {
+  if constexpr (bolt_minor > 0) {
+    if (signature == Signature::Noop) [[unlikely]] {
+      SPDLOG_DEBUG("Received NOOP message");
+      return State::Init;
+    }
+  }
+
+  if (signature != Signature::Init) [[unlikely]] {
+    spdlog::trace("Expected Init signature, but received 0x{:02X}!", utils::UnderlyingCast(signature));
+    return State::Close;
+  }
+
+  auto maybeMetadata = GetMetadataV4(session, marker);
+
+  if (!maybeMetadata) {
+    return State::Close;
+  }
+  if (auto result = AuthenticateUser(session, *maybeMetadata)) {
+    return result.value();
+  }
+
+  return SendSuccessMessage(session);
+}
+}  // namespace details
 
 /**
  * Init state run function.
  * This function runs everything to initialize a Bolt session with the client.
  * @param session the session that should be used for the run.
  */
-template <typename Session>
-State StateInitRun(Session &session) {
+template <typename TSession>
+State StateInitRun(TSession &session) {
   DMG_ASSERT(!session.encoder_buffer_.HasData(), "There should be no data to write in this state");
 
   Marker marker;
@@ -98,72 +198,18 @@ State StateInitRun(Session &session) {
     return State::Close;
   }
 
-  if (UNLIKELY(signature == Signature::Noop && session.version_.major == 4 && session.version_.minor == 1)) {
-    SPDLOG_DEBUG("Received NOOP message");
-    return State::Init;
-  }
-
-  if (UNLIKELY(signature != Signature::Init)) {
-    spdlog::trace("Expected Init signature, but received 0x{:02X}!", utils::UnderlyingCast(signature));
-    return State::Close;
-  }
-
-  auto maybeMetadata =
-      session.version_.major == 1 ? detail::StateInitRunV1(session, marker) : detail::StateInitRunV4(session, marker);
-
-  if (!maybeMetadata) {
-    return State::Close;
-  }
-
-  // Get authentication data.
-  std::string username;
-  std::string password;
-  auto &data = maybeMetadata->ValueMap();
-  if (!data.count("scheme")) {
-    spdlog::warn("The client didn't supply authentication information!");
-    return State::Close;
-  }
-  if (data["scheme"].ValueString() == "basic") {
-    if (!data.count("principal") || !data.count("credentials")) {
-      spdlog::warn("The client didn't supply authentication information!");
-      return State::Close;
+  switch (session.version_.major) {
+    case 1: {
+      return details::StateInitRunV1<TSession>(session, marker, signature);
     }
-    username = data["principal"].ValueString();
-    password = data["credentials"].ValueString();
-  } else if (data["scheme"].ValueString() != "none") {
-    spdlog::warn("Unsupported authentication scheme: {}", data["scheme"].ValueString());
-    return State::Close;
-  }
-
-  // Authenticate the user.
-  if (!session.Authenticate(username, password)) {
-    if (!session.encoder_.MessageFailure(
-            {{"code", "Memgraph.ClientError.Security.Unauthenticated"}, {"message", "Authentication failure"}})) {
-      spdlog::trace("Couldn't send failure message to the client!");
-    }
-    // Throw an exception to indicate to the network stack that the session
-    // should be closed and cleaned up.
-    throw SessionClosedException("The client is not authenticated!");
-  }
-
-  // Return success.
-  {
-    bool success_sent = false;
-    // Neo4j's Java driver 4.1.1+ requires connection_id.
-    // The only usage in the mentioned version is for logging purposes.
-    // Because it's not critical for the regular usage of the driver
-    // we send a hardcoded value for now.
-    std::map<std::string, Value> metadata{{"connection_id", "bolt-1"}};
-    if (auto server_name = session.GetServerNameForInit(); server_name) {
-      metadata.insert({"server", *server_name});
-    }
-    success_sent = session.encoder_.MessageSuccess(metadata);
-    if (!success_sent) {
-      spdlog::trace("Couldn't send success message to the client!");
-      return State::Close;
+    case 4: {
+      if (session.version_.minor > 0) {
+        return details::StateInitRunV4<TSession, 1>(session, marker, signature);
+      }
+      return details::StateInitRunV4<TSession>(session, marker, signature);
     }
   }
-
-  return State::Idle;
+  spdlog::trace("Unsupported bolt version:{}.{})!", session.version_.major, session.version_.minor);
+  return State::Close;
 }
 }  // namespace communication::bolt

src/communication/server.hpp

@@ -89,7 +89,8 @@ class Server final {
     alive_.store(true);
 
     if (!socket_.Bind(endpoint_)) {
-      spdlog::error(utils::MessageWithLink("Cannot bind to socket on endpoint {}.", endpoint_, "https://memgr.ph/socket"));
+      spdlog::error(
+          utils::MessageWithLink("Cannot bind to socket on endpoint {}.", endpoint_, "https://memgr.ph/socket"));
       alive_.store(false);
       return false;
     }
@@ -102,9 +103,8 @@ class Server final {
 
     listener_.Start();
 
-    std::string service_name(service_name_);
-    thread_ = std::thread([this, service_name]() {
-      utils::ThreadSetName(fmt::format("{} server", service_name));
+    thread_ = std::thread([this]() {
+      utils::ThreadSetName(fmt::format("{} server", service_name_));
 
       spdlog::info("{} server is fully armed and operational", service_name_);
       spdlog::info("{} listening on {}", service_name_, socket_.endpoint());

src/integrations/CMakeLists.txt

@@ -1 +1,2 @@
 add_subdirectory(kafka)
+add_subdirectory(pulsar)

src/integrations/constants.hpp

@@ -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 <chrono>
+#include <string>
+
+namespace integrations {
+constexpr int64_t kDefaultCheckBatchLimit{1};
+constexpr std::chrono::milliseconds kDefaultCheckTimeout{30000};
+constexpr std::chrono::milliseconds kMinimumInterval{1};
+constexpr int64_t kMinimumSize{1};
+const std::string kReducted{"<REDUCTED>"};
+}  // namespace integrations

src/integrations/kafka/consumer.cpp

@@ -1,4 +1,4 @@
-// Copyright 2021 Memgraph Ltd.
+// 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
@@ -19,6 +19,8 @@
 
 #include <librdkafka/rdkafkacpp.h>
 #include <spdlog/spdlog.h>
+
+#include "integrations/constants.hpp"
 #include "integrations/kafka/exceptions.hpp"
 #include "utils/exceptions.hpp"
 #include "utils/logging.hpp"
@@ -27,27 +29,19 @@
 
 namespace integrations::kafka {
 
-constexpr std::chrono::milliseconds kDefaultBatchInterval{100};
-constexpr int64_t kDefaultBatchSize = 1000;
-constexpr int64_t kDefaultCheckBatchLimit = 1;
-constexpr std::chrono::milliseconds kDefaultCheckTimeout{30000};
-constexpr std::chrono::milliseconds kMinimumInterval{1};
-constexpr int64_t kMinimumSize{1};
-
 namespace {
 utils::BasicResult<std::string, std::vector<Message>> GetBatch(RdKafka::KafkaConsumer &consumer,
                                                                const ConsumerInfo &info,
                                                                std::atomic<bool> &is_running) {
   std::vector<Message> batch{};
 
-  int64_t batch_size = info.batch_size.value_or(kDefaultBatchSize);
-  batch.reserve(batch_size);
+  batch.reserve(info.batch_size);
 
-  auto remaining_timeout_in_ms = info.batch_interval.value_or(kDefaultBatchInterval).count();
+  auto remaining_timeout_in_ms = info.batch_interval.count();
   auto start = std::chrono::steady_clock::now();
 
   bool run_batch = true;
-  for (int64_t i = 0; remaining_timeout_in_ms > 0 && i < batch_size && is_running.load(); ++i) {
+  for (int64_t i = 0; remaining_timeout_in_ms > 0 && i < info.batch_size && is_running.load(); ++i) {
     std::unique_ptr<RdKafka::Message> msg(consumer.consume(remaining_timeout_in_ms));
     switch (msg->err()) {
       case RdKafka::ERR__TIMED_OUT:
@@ -78,7 +72,7 @@ utils::BasicResult<std::string, std::vector<Message>> GetBatch(RdKafka::KafkaCon
     start = now;
   }
 
-  return {std::move(batch)};
+  return std::move(batch);
 }
 }  // namespace
 
@@ -109,14 +103,19 @@ int64_t Message::Timestamp() const {
   return rd_kafka_message_timestamp(c_message, nullptr);
 }
 
-Consumer::Consumer(const std::string &bootstrap_servers, ConsumerInfo info, ConsumerFunction consumer_function)
-    : info_{std::move(info)}, consumer_function_(std::move(consumer_function)) {
+int64_t Message::Offset() const {
+  const auto *c_message = message_->c_ptr();
+  return c_message->offset;
+}
+
+Consumer::Consumer(ConsumerInfo info, ConsumerFunction consumer_function)
+    : info_{std::move(info)}, consumer_function_(std::move(consumer_function)), cb_(info_.consumer_name) {
   MG_ASSERT(consumer_function_, "Empty consumer function for Kafka consumer");
   // NOLINTNEXTLINE (modernize-use-nullptr)
-  if (info.batch_interval.value_or(kMinimumInterval) < kMinimumInterval) {
+  if (info_.batch_interval < kMinimumInterval) {
     throw ConsumerFailedToInitializeException(info_.consumer_name, "Batch interval has to be positive!");
   }
-  if (info.batch_size.value_or(kMinimumSize) < kMinimumSize) {
+  if (info_.batch_size < kMinimumSize) {
     throw ConsumerFailedToInitializeException(info_.consumer_name, "Batch size has to be positive!");
   }
 
@@ -127,10 +126,26 @@ Consumer::Consumer(const std::string &bootstrap_servers, ConsumerInfo info, Cons
 
   std::string error;
 
+  for (const auto &[key, value] : info_.public_configs) {
+    if (conf->set(key, value, error) != RdKafka::Conf::CONF_OK) {
+      throw SettingCustomConfigFailed(info_.consumer_name, error, key, value);
+    }
+  }
+
+  for (const auto &[key, value] : info_.private_configs) {
+    if (conf->set(key, value, error) != RdKafka::Conf::CONF_OK) {
+      throw SettingCustomConfigFailed(info_.consumer_name, error, key, kReducted);
+    }
+  }
+
   if (conf->set("event_cb", this, error) != RdKafka::Conf::CONF_OK) {
     throw ConsumerFailedToInitializeException(info_.consumer_name, error);
   }
 
+  if (conf->set("rebalance_cb", &cb_, error) != RdKafka::Conf::CONF_OK) {
+    throw ConsumerFailedToInitializeException(info_.consumer_name, error);
+  }
+
   if (conf->set("enable.partition.eof", "false", error) != RdKafka::Conf::CONF_OK) {
     throw ConsumerFailedToInitializeException(info_.consumer_name, error);
   }
@@ -139,7 +154,7 @@ Consumer::Consumer(const std::string &bootstrap_servers, ConsumerInfo info, Cons
     throw ConsumerFailedToInitializeException(info_.consumer_name, error);
   }
 
-  if (conf->set("bootstrap.servers", bootstrap_servers, error) != RdKafka::Conf::CONF_OK) {
+  if (conf->set("bootstrap.servers", info_.bootstrap_servers, error) != RdKafka::Conf::CONF_OK) {
     throw ConsumerFailedToInitializeException(info_.consumer_name, error);
   }
 
@@ -170,7 +185,16 @@ Consumer::Consumer(const std::string &bootstrap_servers, ConsumerInfo info, Cons
                  std::inserter(topic_names_from_metadata, topic_names_from_metadata.begin()),
                  [](const auto topic_metadata) { return topic_metadata->topic(); });
 
+  constexpr size_t max_topic_name_length = 249;
+  constexpr auto is_valid_topic_name = [](const auto c) { return std::isalnum(c) || c == '.' || c == '_' || c == '-'; };
+
   for (const auto &topic_name : info_.topics) {
+    if (topic_name.size() > max_topic_name_length ||
+        std::any_of(topic_name.begin(), topic_name.end(), [&](const auto c) { return !is_valid_topic_name(c); })) {
+      throw ConsumerFailedToInitializeException(info_.consumer_name,
+                                                fmt::format("'{}' is an invalid topic name", topic_name));
+    }
+
     if (!topic_names_from_metadata.contains(topic_name)) {
       throw TopicNotFoundException(info_.consumer_name, topic_name);
     }
@@ -240,10 +264,18 @@ void Consumer::Check(std::optional<std::chrono::milliseconds> timeout, std::opti
   utils::OnScopeExit restore_is_running([this] { is_running_.store(false); });
 
   if (last_assignment_.empty()) {
-    if (const auto err = consumer_->assignment(last_assignment_); err != RdKafka::ERR_NO_ERROR) {
-      spdlog::warn("Saving the commited offset of consumer {} failed: {}", info_.consumer_name, RdKafka::err2str(err));
+    auto throw_consumer_check_failed = [this](const auto err) {
       throw ConsumerCheckFailedException(info_.consumer_name,
                                          fmt::format("Couldn't save commited offsets: '{}'", RdKafka::err2str(err)));
+    };
+    if (const auto err = consumer_->assignment(last_assignment_); err != RdKafka::ERR_NO_ERROR) {
+      spdlog::warn("Saving the assignment of consumer {} failed: {}", info_.consumer_name, RdKafka::err2str(err));
+      throw_consumer_check_failed(err);
+    }
+    if (const auto err = consumer_->position(last_assignment_); err != RdKafka::ERR_NO_ERROR) {
+      spdlog::warn("Saving the position offset assignment of consumer {} failed: {}", info_.consumer_name,
+                   RdKafka::err2str(err));
+      throw_consumer_check_failed(err);
     }
   } else {
     if (const auto err = consumer_->assign(last_assignment_); err != RdKafka::ERR_NO_ERROR) {
@@ -339,7 +371,18 @@ void Consumer::StartConsuming() {
 
       try {
         consumer_function_(batch);
-        if (const auto err = consumer_->commitSync(); err != RdKafka::ERR_NO_ERROR) {
+        std::vector<RdKafka::TopicPartition *> partitions;
+        utils::OnScopeExit clear_partitions([&]() { RdKafka::TopicPartition::destroy(partitions); });
+
+        if (const auto err = consumer_->assignment(partitions); err != RdKafka::ERR_NO_ERROR) {
+          throw ConsumerCheckFailedException(
+              info_.consumer_name, fmt::format("Couldn't get assignment to commit offsets: {}", RdKafka::err2str(err)));
+        }
+        if (const auto err = consumer_->position(partitions); err != RdKafka::ERR_NO_ERROR) {
+          throw ConsumerCheckFailedException(
+              info_.consumer_name, fmt::format("Couldn't get offsets from librdkafka {}", RdKafka::err2str(err)));
+        }
+        if (const auto err = consumer_->commitSync(partitions); err != RdKafka::ERR_NO_ERROR) {
           spdlog::warn("Committing offset of consumer {} failed: {}", info_.consumer_name, RdKafka::err2str(err));
           break;
         }
@@ -358,4 +401,51 @@ void Consumer::StopConsuming() {
   if (thread_.joinable()) thread_.join();
 }
 
+utils::BasicResult<std::string> Consumer::SetConsumerOffsets(int64_t offset) {
+  if (is_running_) {
+    throw ConsumerRunningException(info_.consumer_name);
+  }
+
+  if (offset == -1) {
+    offset = RD_KAFKA_OFFSET_BEGINNING;
+  } else if (offset == -2) {
+    offset = RD_KAFKA_OFFSET_END;
+  }
+
+  cb_.set_offset(offset);
+  if (const auto err = consumer_->subscribe(info_.topics); err != RdKafka::ERR_NO_ERROR) {
+    return fmt::format("Could not set offset of consumer: {}. Error: {}", info_.consumer_name, RdKafka::err2str(err));
+  }
+  return {};
+}
+
+Consumer::ConsumerRebalanceCb::ConsumerRebalanceCb(std::string consumer_name)
+    : consumer_name_(std::move(consumer_name)) {}
+
+void Consumer::ConsumerRebalanceCb::rebalance_cb(RdKafka::KafkaConsumer *consumer, RdKafka::ErrorCode err,
+                                                 std::vector<RdKafka::TopicPartition *> &partitions) {
+  if (err == RdKafka::ERR__REVOKE_PARTITIONS) {
+    consumer->unassign();
+    return;
+  }
+  if (err != RdKafka::ERR__ASSIGN_PARTITIONS) {
+    spdlog::critical("Consumer {} received an unexpected error {}", consumer_name_, RdKafka::err2str(err));
+    return;
+  }
+  if (offset_) {
+    for (auto &partition : partitions) {
+      partition->set_offset(*offset_);
+    }
+    offset_.reset();
+  }
+  auto maybe_error = consumer->assign(partitions);
+  if (maybe_error != RdKafka::ErrorCode::ERR_NO_ERROR) {
+    spdlog::warn("Assigning offset of consumer {} failed: {}", consumer_name_, RdKafka::err2str(maybe_error));
+  }
+  maybe_error = consumer->commitSync(partitions);
+  if (maybe_error != RdKafka::ErrorCode::ERR_NO_ERROR) {
+    spdlog::warn("Commiting offsets of consumer {} failed: {}", consumer_name_, RdKafka::err2str(maybe_error));
+  }
+}
+void Consumer::ConsumerRebalanceCb::set_offset(int64_t offset) { offset_ = offset; }
 }  // namespace integrations::kafka

src/integrations/kafka/consumer.hpp

@@ -1,4 +1,4 @@
-// Copyright 2021 Memgraph Ltd.
+// 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
@@ -17,6 +17,7 @@
 #include <memory>
 #include <optional>
 #include <span>
+#include <string>
 #include <thread>
 #include <utility>
 #include <vector>
@@ -68,6 +69,9 @@ class Message final {
   /// can be implemented knowing that.
   int64_t Timestamp() const;
 
+  /// Returns the offset of the message
+  int64_t Offset() const;
+
  private:
   std::unique_ptr<RdKafka::Message> message_;
 };
@@ -79,8 +83,11 @@ struct ConsumerInfo {
   std::string consumer_name;
   std::vector<std::string> topics;
   std::string consumer_group;
-  std::optional<std::chrono::milliseconds> batch_interval;
-  std::optional<int64_t> batch_size;
+  std::string bootstrap_servers;
+  std::chrono::milliseconds batch_interval;
+  int64_t batch_size;
+  std::unordered_map<std::string, std::string> public_configs;
+  std::unordered_map<std::string, std::string> private_configs;
 };
 
 /// Memgraphs Kafka consumer wrapper.
@@ -93,7 +100,7 @@ class Consumer final : public RdKafka::EventCb {
   ///
   /// @throws ConsumerFailedToInitializeException if the consumer can't connect
   ///         to the Kafka endpoint.
-  Consumer(const std::string &bootstrap_servers, ConsumerInfo info, ConsumerFunction consumer_function);
+  Consumer(ConsumerInfo info, ConsumerFunction consumer_function);
   ~Consumer() override;
 
   Consumer(const Consumer &other) = delete;
@@ -137,6 +144,13 @@ class Consumer final : public RdKafka::EventCb {
   /// Returns true if the consumer is actively consuming messages.
   bool IsRunning() const;
 
+  /// Sets the consumer's offset.
+  ///
+  /// This function returns the empty string on success or an error message otherwise.
+  ///
+  /// @param offset: the offset to set.
+  [[nodiscard]] utils::BasicResult<std::string> SetConsumerOffsets(int64_t offset);
+
   const ConsumerInfo &Info() const;
 
  private:
@@ -146,6 +160,20 @@ class Consumer final : public RdKafka::EventCb {
 
   void StopConsuming();
 
+  class ConsumerRebalanceCb : public RdKafka::RebalanceCb {
+   public:
+    ConsumerRebalanceCb(std::string consumer_name);
+
+    void rebalance_cb(RdKafka::KafkaConsumer *consumer, RdKafka::ErrorCode err,
+                      std::vector<RdKafka::TopicPartition *> &partitions) override final;
+
+    void set_offset(int64_t offset);
+
+   private:
+    std::optional<int64_t> offset_;
+    std::string consumer_name_;
+  };
+
   ConsumerInfo info_;
   ConsumerFunction consumer_function_;
   mutable std::atomic<bool> is_running_{false};
@@ -153,5 +181,6 @@ class Consumer final : public RdKafka::EventCb {
   std::optional<int64_t> limit_batches_{std::nullopt};
   std::unique_ptr<RdKafka::KafkaConsumer, std::function<void(RdKafka::KafkaConsumer *)>> consumer_;
   std::thread thread_;
+  ConsumerRebalanceCb cb_;
 };
 }  // namespace integrations::kafka

src/integrations/kafka/exceptions.hpp

@@ -1,4 +1,4 @@
-// Copyright 2021 Memgraph Ltd.
+// 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
@@ -11,7 +11,7 @@
 
 #pragma once
 
-#include <string>
+#include <string_view>
 
 #include "utils/exceptions.hpp"
 
@@ -22,37 +22,46 @@ class KafkaStreamException : public utils::BasicException {
 
 class ConsumerFailedToInitializeException : public KafkaStreamException {
  public:
-  ConsumerFailedToInitializeException(const std::string &consumer_name, const std::string &error)
+  ConsumerFailedToInitializeException(std::string_view consumer_name, std::string_view error)
       : KafkaStreamException("Failed to initialize Kafka consumer {} : {}", consumer_name, error) {}
 };
 
+class SettingCustomConfigFailed : public ConsumerFailedToInitializeException {
+ public:
+  SettingCustomConfigFailed(std::string_view consumer_name, std::string_view error, std::string_view key,
+                            std::string_view value)
+      : ConsumerFailedToInitializeException(
+            consumer_name,
+            fmt::format(R"(failed to set custom config ("{}": "{}"), because of error {})", key, value, error)) {}
+};
+
 class ConsumerRunningException : public KafkaStreamException {
  public:
-  explicit ConsumerRunningException(const std::string &consumer_name)
+  explicit ConsumerRunningException(std::string_view consumer_name)
       : KafkaStreamException("Kafka consumer {} is already running", consumer_name) {}
 };
 
 class ConsumerStoppedException : public KafkaStreamException {
  public:
-  explicit ConsumerStoppedException(const std::string &consumer_name)
+  explicit ConsumerStoppedException(std::string_view consumer_name)
       : KafkaStreamException("Kafka consumer {} is already stopped", consumer_name) {}
 };
 
 class ConsumerCheckFailedException : public KafkaStreamException {
  public:
-  explicit ConsumerCheckFailedException(const std::string &consumer_name, const std::string &error)
+  explicit ConsumerCheckFailedException(std::string_view consumer_name, std::string_view error)
       : KafkaStreamException("Kafka consumer {} check failed: {}", consumer_name, error) {}
 };
 
 class ConsumerStartFailedException : public KafkaStreamException {
  public:
-  explicit ConsumerStartFailedException(const std::string &consumer_name, const std::string &error)
+  explicit ConsumerStartFailedException(std::string_view consumer_name, std::string_view error)
       : KafkaStreamException("Starting Kafka consumer {} failed: {}", consumer_name, error) {}
 };
 
 class TopicNotFoundException : public KafkaStreamException {
  public:
-  TopicNotFoundException(const std::string &consumer_name, const std::string &topic_name)
+  TopicNotFoundException(std::string_view consumer_name, std::string_view topic_name)
       : KafkaStreamException("Kafka consumer {} cannot find topic {}", consumer_name, topic_name) {}
 };
 }  // namespace integrations::kafka

src/integrations/pulsar/CMakeLists.txt

@@ -0,0 +1,8 @@
+set(integrations_pulsar_src_files
+    consumer.cpp
+)
+
+find_package(CURL REQUIRED)
+
+add_library(mg-integrations-pulsar STATIC ${integrations_pulsar_src_files})
+target_link_libraries(mg-integrations-pulsar mg-utils pulsar Threads::Threads ${CURL_LIBRARIES})

src/integrations/pulsar/consumer.cpp

@@ -0,0 +1,287 @@
+// 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.
+
+#include "integrations/pulsar/consumer.hpp"
+
+#include <fmt/format.h>
+#include <pulsar/Client.h>
+#include <pulsar/InitialPosition.h>
+
+#include <chrono>
+#include <thread>
+
+#include "integrations/constants.hpp"
+#include "integrations/pulsar/exceptions.hpp"
+#include "utils/concepts.hpp"
+#include "utils/logging.hpp"
+#include "utils/on_scope_exit.hpp"
+#include "utils/result.hpp"
+#include "utils/thread.hpp"
+
+namespace integrations::pulsar {
+
+namespace {
+
+template <typename T>
+concept PulsarConsumer = utils::SameAsAnyOf<T, pulsar_client::Consumer, pulsar_client::Reader>;
+
+pulsar_client::Result ConsumeMessage(pulsar_client::Consumer &consumer, pulsar_client::Message &message,
+                                     int remaining_timeout_in_ms) {
+  return consumer.receive(message, remaining_timeout_in_ms);
+}
+
+pulsar_client::Result ConsumeMessage(pulsar_client::Reader &reader, pulsar_client::Message &message,
+                                     int remaining_timeout_in_ms) {
+  return reader.readNext(message, remaining_timeout_in_ms);
+}
+
+template <PulsarConsumer TConsumer>
+utils::BasicResult<std::string, std::vector<Message>> GetBatch(TConsumer &consumer, const ConsumerInfo &info,
+                                                               std::atomic<bool> &is_running,
+                                                               const pulsar_client::MessageId &last_message_id) {
+  std::vector<Message> batch{};
+
+  batch.reserve(info.batch_size);
+
+  auto remaining_timeout_in_ms = info.batch_interval.count();
+  auto start = std::chrono::steady_clock::now();
+
+  while (remaining_timeout_in_ms > 0 && batch.size() < info.batch_size && is_running) {
+    pulsar_client::Message message;
+    const auto result = ConsumeMessage(consumer, message, remaining_timeout_in_ms);
+    switch (result) {
+      case pulsar_client::Result::ResultTimeout:
+        return std::move(batch);
+      case pulsar_client::Result::ResultOk:
+        if (message.getMessageId() != last_message_id) {
+          batch.emplace_back(Message{std::move(message)});
+        }
+        break;
+      default:
+        spdlog::warn(fmt::format("Unexpected error while consuming message from consumer {}, error: {}",
+                                 info.consumer_name, result));
+        return {pulsar_client::strResult(result)};
+    }
+
+    auto now = std::chrono::steady_clock::now();
+    auto took = std::chrono::duration_cast<std::chrono::milliseconds>(now - start);
+    remaining_timeout_in_ms = remaining_timeout_in_ms - took.count();
+    start = now;
+  }
+
+  return std::move(batch);
+}
+
+class SpdlogLogger : public pulsar_client::Logger {
+  bool isEnabled(Level /*level*/) override { return spdlog::should_log(spdlog::level::trace); }
+
+  void log(Level /*level*/, int /*line*/, const std::string &message) override {
+    spdlog::trace("[Pulsar] {}", message);
+  }
+};
+
+class SpdlogLoggerFactory : public pulsar_client::LoggerFactory {
+  pulsar_client::Logger *getLogger(const std::string & /*file_name*/) override { return new SpdlogLogger; }
+};
+
+pulsar_client::Client CreateClient(const std::string &service_url) {
+  pulsar_client::ClientConfiguration conf;
+  conf.setLogger(new SpdlogLoggerFactory);
+  return {service_url, conf};
+}
+}  // namespace
+
+Message::Message(pulsar_client::Message &&message) : message_{std::move(message)} {}
+
+std::span<const char> Message::Payload() const {
+  return {static_cast<const char *>(message_.getData()), message_.getLength()};
+}
+
+std::string_view Message::TopicName() const { return message_.getTopicName(); }
+
+Consumer::Consumer(ConsumerInfo info, ConsumerFunction consumer_function)
+    : info_{std::move(info)},
+      client_{CreateClient(info_.service_url)},
+      consumer_function_{std::move(consumer_function)} {
+  pulsar_client::ConsumerConfiguration config;
+  config.setSubscriptionInitialPosition(pulsar_client::InitialPositionLatest);
+  config.setConsumerType(pulsar_client::ConsumerType::ConsumerExclusive);
+  if (pulsar_client::Result result = client_.subscribe(info_.topics, info_.consumer_name, config, consumer_);
+      result != pulsar_client::ResultOk) {
+    throw ConsumerFailedToInitializeException(info_.consumer_name, pulsar_client::strResult(result));
+  }
+}
+Consumer::~Consumer() {
+  StopIfRunning();
+  consumer_.close();
+  client_.close();
+}
+
+bool Consumer::IsRunning() const { return is_running_; }
+
+const ConsumerInfo &Consumer::Info() const { return info_; }
+
+void Consumer::Start() {
+  if (is_running_) {
+    throw ConsumerRunningException(info_.consumer_name);
+  }
+
+  StartConsuming();
+}
+
+void Consumer::Stop() {
+  if (!is_running_) {
+    throw ConsumerStoppedException(info_.consumer_name);
+  }
+  StopConsuming();
+}
+
+void Consumer::StopIfRunning() {
+  if (is_running_) {
+    StopConsuming();
+  }
+
+  if (thread_.joinable()) {
+    thread_.join();
+  }
+}
+
+void Consumer::Check(std::optional<std::chrono::milliseconds> timeout, std::optional<int64_t> limit_batches,
+                     const ConsumerFunction &check_consumer_function) const {
+  // NOLINTNEXTLINE (modernize-use-nullptr)
+  if (timeout.value_or(kMinimumInterval) < kMinimumInterval) {
+    throw ConsumerCheckFailedException(info_.consumer_name, "Timeout has to be positive!");
+  }
+  if (limit_batches.value_or(kMinimumSize) < kMinimumSize) {
+    throw ConsumerCheckFailedException(info_.consumer_name, "Batch limit has to be positive!");
+  }
+  // The implementation of this function is questionable: it is const qualified, though it changes the inner state of
+  // PulsarConsumer. Though it changes the inner state, it saves the current assignment for future Check/Start calls to
+  // restore the current state, so the changes made by this function shouldn't be visible for the users of the class. It
+  // also passes a non const reference of PulsarConsumer to GetBatch function. That means the object is bitwise const
+  // (PulsarConsumer is stored in unique_ptr) and internally mostly synchronized. Mostly, because as Start/Stop requires
+  // exclusive access to consumer, so we don't have to deal with simultaneous calls to those functions. The only concern
+  // in this function is to prevent executing this function on multiple threads simultaneously.
+  if (is_running_.exchange(true)) {
+    throw ConsumerRunningException(info_.consumer_name);
+  }
+
+  utils::OnScopeExit restore_is_running([this] { is_running_.store(false); });
+
+  const auto num_of_batches = limit_batches.value_or(kDefaultCheckBatchLimit);
+  const auto timeout_to_use = timeout.value_or(kDefaultCheckTimeout);
+  const auto start = std::chrono::steady_clock::now();
+
+  if (info_.topics.size() != 1) {
+    throw ConsumerCheckFailedException(info_.consumer_name, "Check cannot be used for consumers with multiple topics.");
+  }
+
+  std::vector<std::string> partitions;
+  const auto &topic = info_.topics.front();
+  client_.getPartitionsForTopic(topic, partitions);
+  if (partitions.size() > 1) {
+    throw ConsumerCheckFailedException(info_.consumer_name, "Check cannot be used for topics with multiple partitions");
+  }
+  pulsar_client::Reader reader;
+  client_.createReader(topic, last_message_id_, {}, reader);
+  for (int64_t i = 0; i < num_of_batches;) {
+    const auto now = std::chrono::steady_clock::now();
+    // NOLINTNEXTLINE (modernize-use-nullptr)
+    if (now - start >= timeout_to_use) {
+      throw ConsumerCheckFailedException(info_.consumer_name, "Timeout reached");
+    }
+
+    auto maybe_batch = GetBatch(reader, info_, is_running_, last_message_id_);
+
+    if (maybe_batch.HasError()) {
+      throw ConsumerCheckFailedException(info_.consumer_name, maybe_batch.GetError());
+    }
+
+    const auto &batch = maybe_batch.GetValue();
+
+    if (batch.empty()) {
+      continue;
+    }
+    ++i;
+
+    try {
+      check_consumer_function(batch);
+    } catch (const std::exception &e) {
+      spdlog::warn("Pulsar consumer {} check failed with error {}", info_.consumer_name, e.what());
+      throw ConsumerCheckFailedException(info_.consumer_name, e.what());
+    }
+  }
+  reader.close();
+}
+
+void Consumer::StartConsuming() {
+  MG_ASSERT(!is_running_, "Cannot start already running consumer!");
+  if (thread_.joinable()) {
+    thread_.join();
+  }
+
+  is_running_.store(true);
+
+  thread_ = std::thread([this] {
+    constexpr auto kMaxThreadNameSize = utils::GetMaxThreadNameSize();
+    const auto full_thread_name = "Cons#" + info_.consumer_name;
+
+    utils::ThreadSetName(full_thread_name.substr(0, kMaxThreadNameSize));
+
+    while (is_running_) {
+      auto maybe_batch = GetBatch(consumer_, info_, is_running_, last_message_id_);
+
+      if (maybe_batch.HasError()) {
+        spdlog::warn("Error happened in consumer {} while fetching messages: {}!", info_.consumer_name,
+                     maybe_batch.GetError());
+        break;
+      }
+
+      const auto &batch = maybe_batch.GetValue();
+
+      if (batch.empty()) {
+        continue;
+      }
+
+      spdlog::info("Pulsar consumer {} is processing a batch", info_.consumer_name);
+
+      try {
+        consumer_function_(batch);
+
+        if (std::any_of(batch.begin(), batch.end(), [&](const auto &message) {
+              if (const auto result = consumer_.acknowledge(message.message_); result != pulsar_client::ResultOk) {
+                spdlog::warn("Acknowledging a message of consumer {} failed: {}", info_.consumer_name, result);
+                return true;
+              }
+              last_message_id_ = message.message_.getMessageId();
+              return false;
+            })) {
+          break;
+        }
+      } catch (const std::exception &e) {
+        spdlog::warn("Error happened in consumer {} while processing a batch: {}!", info_.consumer_name, e.what());
+        break;
+      }
+
+      spdlog::info("Pulsar consumer {} finished processing", info_.consumer_name);
+    }
+    is_running_.store(false);
+  });
+}
+
+void Consumer::StopConsuming() {
+  is_running_.store(false);
+  if (thread_.joinable()) {
+    thread_.join();
+  }
+}
+
+}  // namespace integrations::pulsar

src/integrations/pulsar/consumer.hpp

@@ -0,0 +1,82 @@
+// 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.
+
+#pragma once
+#include <atomic>
+#include <optional>
+#include <span>
+#include <thread>
+
+#include <pulsar/Client.h>
+
+namespace integrations::pulsar {
+
+namespace pulsar_client = ::pulsar;
+
+class Consumer;
+
+class Message final {
+ public:
+  explicit Message(pulsar_client::Message &&message);
+
+  std::span<const char> Payload() const;
+  std::string_view TopicName() const;
+
+ private:
+  pulsar_client::Message message_;
+
+  friend Consumer;
+};
+
+using ConsumerFunction = std::function<void(const std::vector<Message> &)>;
+
+struct ConsumerInfo {
+  int64_t batch_size;
+  std::chrono::milliseconds batch_interval;
+  std::vector<std::string> topics;
+  std::string consumer_name;
+  std::string service_url;
+};
+
+class Consumer final {
+ public:
+  Consumer(ConsumerInfo info, ConsumerFunction consumer_function);
+  ~Consumer();
+
+  Consumer(const Consumer &) = delete;
+  Consumer(Consumer &&) noexcept = delete;
+  Consumer &operator=(const Consumer &) = delete;
+  Consumer &operator=(Consumer &&) = delete;
+
+  bool IsRunning() const;
+  void Start();
+  void Stop();
+  void StopIfRunning();
+
+  void Check(std::optional<std::chrono::milliseconds> timeout, std::optional<int64_t> limit_batches,
+             const ConsumerFunction &check_consumer_function) const;
+
+  const ConsumerInfo &Info() const;
+
+ private:
+  void StartConsuming();
+  void StopConsuming();
+
+  ConsumerInfo info_;
+  mutable pulsar_client::Client client_;
+  pulsar_client::Consumer consumer_;
+  ConsumerFunction consumer_function_;
+
+  mutable std::atomic<bool> is_running_{false};
+  pulsar_client::MessageId last_message_id_{pulsar_client::MessageId::earliest()};
+  std::thread thread_;
+};
+}  // namespace integrations::pulsar

src/integrations/pulsar/exceptions.hpp

@@ -0,0 +1,58 @@
+// 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.
+
+#pragma once
+
+#include <string>
+
+#include "utils/exceptions.hpp"
+
+namespace integrations::pulsar {
+class PulsarStreamException : public utils::BasicException {
+  using utils::BasicException::BasicException;
+};
+
+class ConsumerFailedToInitializeException : public PulsarStreamException {
+ public:
+  ConsumerFailedToInitializeException(const std::string &consumer_name, const std::string &error)
+      : PulsarStreamException("Failed to initialize Pulsar consumer {} : {}", consumer_name, error) {}
+};
+
+class ConsumerRunningException : public PulsarStreamException {
+ public:
+  explicit ConsumerRunningException(const std::string &consumer_name)
+      : PulsarStreamException("Pulsar consumer {} is already running", consumer_name) {}
+};
+
+class ConsumerStoppedException : public PulsarStreamException {
+ public:
+  explicit ConsumerStoppedException(const std::string &consumer_name)
+      : PulsarStreamException("Pulsar consumer {} is already stopped", consumer_name) {}
+};
+
+class ConsumerCheckFailedException : public PulsarStreamException {
+ public:
+  explicit ConsumerCheckFailedException(const std::string &consumer_name, const std::string &error)
+      : PulsarStreamException("Pulsar consumer {} check failed: {}", consumer_name, error) {}
+};
+
+class ConsumerStartFailedException : public PulsarStreamException {
+ public:
+  explicit ConsumerStartFailedException(const std::string &consumer_name, const std::string &error)
+      : PulsarStreamException("Starting Pulsar consumer {} failed: {}", consumer_name, error) {}
+};
+
+class TopicNotFoundException : public PulsarStreamException {
+ public:
+  TopicNotFoundException(const std::string &consumer_name, const std::string &topic_name)
+      : PulsarStreamException("Pulsar consumer {} cannot find topic {}", consumer_name, topic_name) {}
+};
+}  // namespace integrations::pulsar

src/io/CMakeLists.txt

@@ -4,5 +4,8 @@ set(io_src_files
     network/socket.cpp
     network/utils.cpp)
 
+find_package(fmt REQUIRED)
+find_package(Threads REQUIRED)
+
 add_library(mg-io STATIC ${io_src_files})
-target_link_libraries(mg-io stdc++fs Threads::Threads fmt mg-utils)
+target_link_libraries(mg-io stdc++fs Threads::Threads fmt::fmt mg-utils)

src/kvstore/CMakeLists.txt

@@ -1,6 +1,10 @@
+find_package(gflags REQUIRED)
+find_package(BZip2 REQUIRED)
+find_package(ZLIB REQUIRED)
+
 # STATIC library used to store key-value pairs
 add_library(mg-kvstore STATIC kvstore.cpp)
-target_link_libraries(mg-kvstore stdc++fs mg-utils rocksdb bzip2 zlib gflags)
+target_link_libraries(mg-kvstore stdc++fs mg-utils rocksdb BZip2::BZip2 ZLIB::ZLIB gflags)
 
 # STATIC library for dummy key-value storage
 # add_library(mg-kvstore-dummy STATIC kvstore_dummy.cpp)

src/memgraph.cpp

@@ -186,10 +186,22 @@ DEFINE_bool(telemetry_enabled, false,
             "the database runtime (vertex and edge counts and resource usage) "
             "to allow for easier improvement of the product.");
 
+// Streams flags
+// NOLINTNEXTLINE (cppcoreguidelines-avoid-non-const-global-variables)
+DEFINE_uint32(
+    stream_transaction_conflict_retries, 30,
+    "Number of times to retry when a stream transformation fails to commit because of conflicting transactions");
+// NOLINTNEXTLINE (cppcoreguidelines-avoid-non-const-global-variables)
+DEFINE_uint32(
+    stream_transaction_retry_interval, 500,
+    "Retry interval in milliseconds when a stream transformation fails to commit because of conflicting transactions");
 // NOLINTNEXTLINE (cppcoreguidelines-avoid-non-const-global-variables)
 DEFINE_string(kafka_bootstrap_servers, "",
               "List of default Kafka brokers as a comma separated list of broker host or host:port.");
 
+// NOLINTNEXTLINE (cppcoreguidelines-avoid-non-const-global-variables)
+DEFINE_string(pulsar_service_url, "", "Default URL used while connecting to Pulsar brokers.");
+
 // Audit logging flags.
 #ifdef MG_ENTERPRISE
 DEFINE_bool(audit_enabled, false, "Set to true to enable audit logging.");
@@ -1121,9 +1133,13 @@ int main(int argc, char **argv) {
 
   query::InterpreterContext interpreter_context{
       &db,
-      {.query = {.allow_load_csv = FLAGS_allow_load_csv}, .execution_timeout_sec = FLAGS_query_execution_timeout_sec},
-      FLAGS_data_directory,
-      FLAGS_kafka_bootstrap_servers};
+      {.query = {.allow_load_csv = FLAGS_allow_load_csv},
+       .execution_timeout_sec = FLAGS_query_execution_timeout_sec,
+       .default_kafka_bootstrap_servers = FLAGS_kafka_bootstrap_servers,
+       .default_pulsar_service_url = FLAGS_pulsar_service_url,
+       .stream_transaction_conflict_retries = FLAGS_stream_transaction_conflict_retries,
+       .stream_transaction_retry_interval = std::chrono::milliseconds(FLAGS_stream_transaction_retry_interval)},
+      FLAGS_data_directory};
 #ifdef MG_ENTERPRISE
   SessionData session_data{&db, &interpreter_context, &auth, &audit_log};
 #else
@@ -1133,9 +1149,6 @@ int main(int argc, char **argv) {
   query::procedure::gModuleRegistry.SetModulesDirectory(query_modules_directories);
   query::procedure::gModuleRegistry.UnloadAndLoadModulesFromDirectories();
 
-  // As the Stream transformations are using modules, they have to be restored after the query modules are loaded.
-  interpreter_context.streams.RestoreStreams();
-
   AuthQueryHandler auth_handler(&auth, FLAGS_auth_user_or_role_name_regex);
   AuthChecker auth_checker{&auth};
   interpreter_context.auth = &auth_handler;
@@ -1151,6 +1164,9 @@ int main(int argc, char **argv) {
                                                       interpreter_context.auth_checker);
   }
 
+  // As the Stream transformations are using modules, they have to be restored after the query modules are loaded.
+  interpreter_context.streams.RestoreStreams();
+
   ServerContext context;
   std::string service_name = "Bolt";
   if (!FLAGS_bolt_key_file.empty() && !FLAGS_bolt_cert_file.empty()) {

src/memory/CMakeLists.txt

@@ -2,9 +2,11 @@ set(memory_src_files
     new_delete.cpp
     memory_control.cpp)
 
+find_package(Jemalloc REQUIRED)
+
 add_library(mg-memory STATIC ${memory_src_files})
 target_link_libraries(mg-memory mg-utils fmt)
 
 if (ENABLE_JEMALLOC)
-  target_link_libraries(mg-memory jemalloc)
+  target_link_libraries(mg-memory Jemalloc::Jemalloc)
 endif()

src/memory/new_delete.cpp

@@ -19,13 +19,12 @@
 #include <cstdlib>
 #endif
 
-#include "utils/likely.hpp"
 #include "utils/memory_tracker.hpp"
 
 namespace {
 void *newImpl(const std::size_t size) {
   auto *ptr = malloc(size);
-  if (LIKELY(ptr != nullptr)) {
+  if (ptr != nullptr) [[likely]] {
     return ptr;
   }
 
@@ -34,7 +33,7 @@ void *newImpl(const std::size_t size) {
 
 void *newImpl(const std::size_t size, const std::align_val_t align) {
   auto *ptr = aligned_alloc(static_cast<std::size_t>(align), size);
-  if (LIKELY(ptr != nullptr)) {
+  if (ptr != nullptr) [[likely]] {
     return ptr;
   }
 
@@ -47,14 +46,22 @@ void *newNoExcept(const std::size_t size, const std::align_val_t align) noexcept
 }
 
 #if USE_JEMALLOC
-void deleteImpl(void *ptr) noexcept { dallocx(ptr, 0); }
+void deleteImpl(void *ptr) noexcept {
+  if (ptr == nullptr) [[unlikely]] {
+    return;
+  }
+  dallocx(ptr, 0);
+}
 
 void deleteImpl(void *ptr, const std::align_val_t align) noexcept {
+  if (ptr == nullptr) [[unlikely]] {
+    return;
+  }
   dallocx(ptr, MALLOCX_ALIGN(align));  // NOLINT(hicpp-signed-bitwise)
 }
 
 void deleteSized(void *ptr, const std::size_t size) noexcept {
-  if (UNLIKELY(ptr == nullptr)) {
+  if (ptr == nullptr) [[unlikely]] {
     return;
   }
 
@@ -62,7 +69,7 @@ void deleteSized(void *ptr, const std::size_t size) noexcept {
 }
 
 void deleteSized(void *ptr, const std::size_t size, const std::align_val_t align) noexcept {
-  if (UNLIKELY(ptr == nullptr)) {
+  if (ptr == nullptr) [[unlikely]] {
     return;
   }
 
@@ -81,7 +88,7 @@ void deleteSized(void *ptr, const std::size_t /*unused*/, const std::align_val_t
 
 void TrackMemory(std::size_t size) {
 #if USE_JEMALLOC
-  if (LIKELY(size != 0)) {
+  if (size != 0) [[likely]] {
     size = nallocx(size, 0);
   }
 #endif
@@ -90,7 +97,7 @@ void TrackMemory(std::size_t size) {
 
 void TrackMemory(std::size_t size, const std::align_val_t align) {
 #if USE_JEMALLOC
-  if (LIKELY(size != 0)) {
+  if (size != 0) [[likely]] {
     size = nallocx(size, MALLOCX_ALIGN(align));  // NOLINT(hicpp-signed-bitwise)
   }
 #endif
@@ -120,7 +127,7 @@ bool TrackMemoryNoExcept(const std::size_t size, const std::align_val_t align) {
 void UntrackMemory([[maybe_unused]] void *ptr, [[maybe_unused]] std::size_t size = 0) noexcept {
   try {
 #if USE_JEMALLOC
-    if (LIKELY(ptr != nullptr)) {
+    if (ptr != nullptr) [[likely]] {
       utils::total_memory_tracker.Free(sallocx(ptr, 0));
     }
 #else
@@ -138,7 +145,7 @@ void UntrackMemory([[maybe_unused]] void *ptr, [[maybe_unused]] std::size_t size
 void UntrackMemory(void *ptr, const std::align_val_t align, [[maybe_unused]] std::size_t size = 0) noexcept {
   try {
 #if USE_JEMALLOC
-    if (LIKELY(ptr != nullptr)) {
+    if (ptr != nullptr) [[likely]] {
       utils::total_memory_tracker.Free(sallocx(ptr, MALLOCX_ALIGN(align)));  // NOLINT(hicpp-signed-bitwise)
     }
 #else
@@ -176,28 +183,28 @@ void *operator new[](const std::size_t size, const std::align_val_t align) {
 }
 
 void *operator new(const std::size_t size, const std::nothrow_t & /*unused*/) noexcept {
-  if (LIKELY(TrackMemoryNoExcept(size))) {
+  if (TrackMemoryNoExcept(size)) [[likely]] {
     return newNoExcept(size);
   }
   return nullptr;
 }
 
 void *operator new[](const std::size_t size, const std::nothrow_t & /*unused*/) noexcept {
-  if (LIKELY(TrackMemoryNoExcept(size))) {
+  if (TrackMemoryNoExcept(size)) [[likely]] {
     return newNoExcept(size);
   }
   return nullptr;
 }
 
 void *operator new(const std::size_t size, const std::align_val_t align, const std::nothrow_t & /*unused*/) noexcept {
-  if (LIKELY(TrackMemoryNoExcept(size, align))) {
+  if (TrackMemoryNoExcept(size, align)) [[likely]] {
     return newNoExcept(size, align);
   }
   return nullptr;
 }
 
 void *operator new[](const std::size_t size, const std::align_val_t align, const std::nothrow_t & /*unused*/) noexcept {
-  if (LIKELY(TrackMemoryNoExcept(size, align))) {
+  if (TrackMemoryNoExcept(size, align)) [[likely]] {
     return newNoExcept(size, align);
   }
   return nullptr;

src/query/CMakeLists.txt

@@ -20,6 +20,7 @@ set(mg_query_sources
     interpret/awesome_memgraph_functions.cpp
     interpret/eval.cpp
     interpreter.cpp
+    metadata.cpp
     plan/operator.cpp
     plan/preprocess.cpp
     plan/pretty_print.cpp
@@ -29,19 +30,27 @@ set(mg_query_sources
     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
-    streams.cpp
+    stream/streams.cpp
+    stream/sources.cpp
+    stream/common.cpp
     trigger.cpp
     trigger_context.cpp
     typed_value.cpp)
 
+find_package(Boost REQUIRED)
+
 add_library(mg-query STATIC ${mg_query_sources})
 add_dependencies(mg-query generate_lcp_query)
 target_include_directories(mg-query PUBLIC ${CMAKE_SOURCE_DIR}/include)
-target_link_libraries(mg-query dl cppitertools)
-target_link_libraries(mg-query mg-integrations-kafka mg-storage-v2 mg-utils mg-kvstore mg-memory)
+target_link_libraries(mg-query dl cppitertools Boost::headers)
+target_link_libraries(mg-query mg-integrations-pulsar mg-integrations-kafka mg-storage-v2 mg-license mg-utils mg-kvstore mg-memory)
+if(NOT "${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()

src/query/common.cpp

@@ -42,15 +42,28 @@ bool TypedValueCompare(const TypedValue &a, const TypedValue &b) {
       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());
-    default:
-      LOG_FATAL("Unhandled comparison for types");
+    case TypedValue::Type::Null:
+      LOG_FATAL("Invalid type");
   }
 }
 

src/query/common.hpp

@@ -33,10 +33,6 @@ namespace impl {
 bool TypedValueCompare(const TypedValue &a, const TypedValue &b);
 }  // namespace impl
 
-constexpr inline std::string_view kSerializationErrorMessage{
-    "Cannot resolve conflicting transactions. You can retry this transaction when the conflicting transaction is "
-    "finished."};
-
 /// Custom Comparator type for comparing vectors of TypedValues.
 ///
 /// Does lexicographical ordering of elements based on the above
@@ -95,7 +91,7 @@ storage::PropertyValue PropsSetChecked(T *record, const storage::PropertyId &key
     if (maybe_old_value.HasError()) {
       switch (maybe_old_value.GetError()) {
         case storage::Error::SERIALIZATION_ERROR:
-          throw QueryRuntimeException(kSerializationErrorMessage);
+          throw TransactionSerializationException();
         case storage::Error::DELETED_OBJECT:
           throw QueryRuntimeException("Trying to set properties on a deleted object.");
         case storage::Error::PROPERTIES_DISABLED:

src/query/config.hpp

@@ -10,6 +10,8 @@
 // licenses/APL.txt.
 
 #pragma once
+#include <chrono>
+#include <string>
 
 namespace query {
 struct InterpreterConfig {
@@ -19,5 +21,10 @@ struct InterpreterConfig {
 
   // The default execution timeout is 10 minutes.
   double execution_timeout_sec{600.0};
+
+  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 query

src/query/context.hpp

@@ -15,6 +15,7 @@
 
 #include "query/common.hpp"
 #include "query/frontend/semantic/symbol_table.hpp"
+#include "query/metadata.hpp"
 #include "query/parameters.hpp"
 #include "query/plan/profile.hpp"
 #include "query/trigger.hpp"
@@ -68,6 +69,7 @@ struct ExecutionContext {
   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;
 };

src/query/exceptions.hpp

@@ -130,6 +130,18 @@ class ExplicitTransactionUsageException : public QueryRuntimeException {
   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()

src/query/frontend/ast/ast.lcp

@@ -54,9 +54,7 @@ cpp<#
     size_t size = 0;
     slk::Load(&size, reader);
     self->${member}.resize(size);
-    for (size_t i = 0;
-         i < size;
-         ++i) {
+    for (size_t i = 0; i < size; ++i) {
       self->${member}[i] = query::LoadAstPointer<query::${type}>(storage, reader);
     }
     cpp<#))
@@ -75,9 +73,7 @@ cpp<#
   #>cpp
   size_t size = 0;
   slk::Load(&size, reader);
-  for (size_t i = 0;
-       i < size;
-       ++i) {
+  for (size_t i = 0; i < size; ++i) {
     query::PropertyIx key;
     slk::Load(&key, reader, storage);
     auto *value = query::LoadAstPointer<query::Expression>(storage, reader);
@@ -93,6 +89,34 @@ cpp<#
   }
   cpp<#)
 
+(defun slk-save-expression-map (member)
+  #>cpp
+  size_t size = self.${member}.size();
+  slk::Save(size, builder);
+  for (const auto &entry : self.${member}) {
+    query::SaveAstPointer(entry.first, builder);
+    query::SaveAstPointer(entry.second, builder);
+  }
+  cpp<#)
+
+(defun slk-load-expression-map (member)
+  #>cpp
+  size_t size = 0;
+  slk::Load(&size, reader);
+  for (size_t i = 0; i < size; ++i) {
+    auto *key = query::LoadAstPointer<query::Expression>(storage, reader);
+    auto *value = query::LoadAstPointer<query::Expression>(storage, reader);
+    self->${member}.emplace(key, value);
+  }
+  cpp<#)
+
+(defun clone-expression-map (source dest)
+  #>cpp
+  for (const auto &[key, value] : ${source}) {
+    ${dest}[key->Clone(storage)] = value->Clone(storage);
+  }
+  cpp<#)
+
 (defun slk-load-name-ix (name-type)
   (lambda (member)
     #>cpp
@@ -1819,9 +1843,7 @@ cpp<#
                          size_t size = 0;
                          slk::Load(&size, reader);
                          self->${member}.resize(size);
-                         for (size_t i = 0;
-                              i < size;
-                              ++i) {
+                         for (size_t i = 0; i < size; ++i) {
                            slk::Load(&self->${member}[i], reader, storage);
                          }
                          cpp<#)
@@ -2498,32 +2520,68 @@ cpp<#
   (:serialize (:slk))
   (:clone))
 
+(defun clone-variant-topic-names (source destination)
+  #>cpp
+    if (auto *topic_expression = std::get_if<Expression*>(&${source})) {
+      if (*topic_expression == nullptr) {
+        ${destination} = nullptr;
+      } else {
+        ${destination} = (*topic_expression)->Clone(storage);
+      }
+    } else {
+      ${destination} = std::get<std::vector<std::string>>(${source});
+    }
+  cpp<#)
+
 (lcp:define-class stream-query (query)
   ((action "Action" :scope :public)
+   (type "Type" :scope :public)
    (stream_name "std::string" :scope :public)
-   (topic_names "std::vector<std::string>" :scope :public)
-   (transform_name "std::string" :scope :public)
-   (consumer_group "std::string" :scope :public)
-   (batch_interval "Expression *" :initval "nullptr" :scope :public
-             :slk-save #'slk-save-ast-pointer
-             :slk-load (slk-load-ast-pointer "Expression"))
-   (batch_size "Expression *" :initval "nullptr" :scope :public
-             :slk-save #'slk-save-ast-pointer
-             :slk-load (slk-load-ast-pointer "Expression"))
+
    (batch_limit "Expression *" :initval "nullptr" :scope :public
              :slk-save #'slk-save-ast-pointer
              :slk-load (slk-load-ast-pointer "Expression"))
    (timeout "Expression *" :initval "nullptr" :scope :public
              :slk-save #'slk-save-ast-pointer
              :slk-load (slk-load-ast-pointer "Expression"))
+
+   (transform_name "std::string" :scope :public)
+   (batch_interval "Expression *" :initval "nullptr" :scope :public
+             :slk-save #'slk-save-ast-pointer
+             :slk-load (slk-load-ast-pointer "Expression"))
+   (batch_size "Expression *" :initval "nullptr" :scope :public
+             :slk-save #'slk-save-ast-pointer
+             :slk-load (slk-load-ast-pointer "Expression"))
+
+   (topic_names "std::variant<Expression*, std::vector<std::string>>" :initval "nullptr"
+             :clone #'clone-variant-topic-names
+             :scope :public)
+   (consumer_group "std::string" :scope :public)
    (bootstrap_servers "Expression *" :initval "nullptr" :scope :public
              :slk-save #'slk-save-ast-pointer
-             :slk-load (slk-load-ast-pointer "Expression")))
+             :slk-load (slk-load-ast-pointer "Expression"))
+
+   (service_url "Expression *" :initval "nullptr" :scope :public
+             :slk-save #'slk-save-ast-pointer
+             :slk-load (slk-load-ast-pointer "Expression"))
+
+   (configs "std::unordered_map<Expression *, Expression *>" :scope :public
+             :slk-save #'slk-save-expression-map
+             :slk-load #'slk-load-expression-map
+             :clone #'clone-expression-map)
+
+   (credentials "std::unordered_map<Expression *, Expression *>" :scope :public
+             :slk-save #'slk-save-expression-map
+             :slk-load #'slk-load-expression-map
+             :clone #'clone-expression-map))
 
   (:public
     (lcp:define-enum action
         (create-stream drop-stream start-stream stop-stream start-all-streams stop-all-streams show-streams check-stream)
       (:serialize))
+    (lcp:define-enum type
+        (kafka pulsar)
+      (:serialize))
     #>cpp
     StreamQuery() = default;
 

src/query/frontend/ast/cypher_main_visitor.cpp

@@ -1,4 +1,4 @@
-// Copyright 2021 Memgraph Ltd.
+// 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
@@ -18,6 +18,7 @@
 // of the same name, EOF.
 // This hides the definition of the macro which causes
 // the compilation to fail.
+#include "query/frontend/ast/ast.hpp"
 #include "query/frontend/ast/ast_visitor.hpp"
 #include "query/procedure/module.hpp"
 //////////////////////////////////////////////////////
@@ -31,14 +32,18 @@
 #include <limits>
 #include <string>
 #include <tuple>
+#include <type_traits>
 #include <unordered_map>
 #include <utility>
 #include <variant>
 #include <vector>
 
+#include <boost/preprocessor/cat.hpp>
+
 #include "query/exceptions.hpp"
 #include "query/frontend/parsing.hpp"
 #include "query/interpret/awesome_memgraph_functions.hpp"
+#include "query/stream/common.hpp"
 #include "utils/exceptions.hpp"
 #include "utils/logging.hpp"
 #include "utils/string.hpp"
@@ -492,45 +497,281 @@ antlrcpp::Any CypherMainVisitor::visitStreamQuery(MemgraphCypher::StreamQueryCon
 }
 
 antlrcpp::Any CypherMainVisitor::visitCreateStream(MemgraphCypher::CreateStreamContext *ctx) {
+  MG_ASSERT(ctx->children.size() == 1, "CreateStreamQuery should have exactly one child!");
+  auto *stream_query = ctx->children[0]->accept(this).as<StreamQuery *>();
+  query_ = stream_query;
+  return stream_query;
+}
+
+namespace {
+std::vector<std::string> TopicNamesFromSymbols(
+    antlr4::tree::ParseTreeVisitor &visitor,
+    const std::vector<MemgraphCypher::SymbolicNameWithDotsAndMinusContext *> &topic_name_symbols) {
+  MG_ASSERT(!topic_name_symbols.empty());
+  std::vector<std::string> topic_names;
+  topic_names.reserve(topic_name_symbols.size());
+  std::transform(topic_name_symbols.begin(), topic_name_symbols.end(), std::back_inserter(topic_names),
+                 [&visitor](auto *topic_name) { return JoinSymbolicNamesWithDotsAndMinus(visitor, *topic_name); });
+  return topic_names;
+}
+
+template <typename T>
+concept EnumUint8 = std::is_enum_v<T> && std::same_as<uint8_t, std::underlying_type_t<T>>;
+
+template <bool required, typename... ValueTypes>
+void MapConfig(auto &memory, const EnumUint8 auto &enum_key, auto &destination) {
+  const auto key = static_cast<uint8_t>(enum_key);
+  if (!memory.contains(key)) {
+    if constexpr (required) {
+      throw SemanticException("Config {} is required.", ToString(enum_key));
+    } else {
+      return;
+    }
+  }
+
+  std::visit(
+      [&]<typename T>(T &&value) {
+        using ValueType = std::decay_t<T>;
+        if constexpr (utils::SameAsAnyOf<ValueType, ValueTypes...>) {
+          destination = std::forward<T>(value);
+        } else {
+          LOG_FATAL("Invalid type mapped");
+        }
+      },
+      std::move(memory[key]));
+  memory.erase(key);
+}
+
+enum class CommonStreamConfigKey : uint8_t { TRANSFORM, BATCH_INTERVAL, BATCH_SIZE, END };
+
+std::string_view ToString(const CommonStreamConfigKey key) {
+  switch (key) {
+    case CommonStreamConfigKey::TRANSFORM:
+      return "TRANSFORM";
+    case CommonStreamConfigKey::BATCH_INTERVAL:
+      return "BATCH_INTERVAL";
+    case CommonStreamConfigKey::BATCH_SIZE:
+      return "BATCH_SIZE";
+    case CommonStreamConfigKey::END:
+      LOG_FATAL("Invalid config key used");
+  }
+}
+
+// NOLINTNEXTLINE(cppcoreguidelines-macro-usage)
+#define GENERATE_STREAM_CONFIG_KEY_ENUM(stream, first_config, ...)   \
+  enum class BOOST_PP_CAT(stream, ConfigKey) : uint8_t {             \
+    first_config = static_cast<uint8_t>(CommonStreamConfigKey::END), \
+    __VA_ARGS__                                                      \
+  };
+
+GENERATE_STREAM_CONFIG_KEY_ENUM(Kafka, TOPICS, CONSUMER_GROUP, BOOTSTRAP_SERVERS, CONFIGS, CREDENTIALS);
+
+std::string_view ToString(const KafkaConfigKey key) {
+  switch (key) {
+    case KafkaConfigKey::TOPICS:
+      return "TOPICS";
+    case KafkaConfigKey::CONSUMER_GROUP:
+      return "CONSUMER_GROUP";
+    case KafkaConfigKey::BOOTSTRAP_SERVERS:
+      return "BOOTSTRAP_SERVERS";
+    case KafkaConfigKey::CONFIGS:
+      return "CONFIGS";
+    case KafkaConfigKey::CREDENTIALS:
+      return "CREDENTIALS";
+  }
+}
+
+void MapCommonStreamConfigs(auto &memory, StreamQuery &stream_query) {
+  MapConfig<true, std::string>(memory, CommonStreamConfigKey::TRANSFORM, stream_query.transform_name_);
+  MapConfig<false, Expression *>(memory, CommonStreamConfigKey::BATCH_INTERVAL, stream_query.batch_interval_);
+  MapConfig<false, Expression *>(memory, CommonStreamConfigKey::BATCH_SIZE, stream_query.batch_size_);
+}
+}  // namespace
+
+antlrcpp::Any CypherMainVisitor::visitConfigKeyValuePair(MemgraphCypher::ConfigKeyValuePairContext *ctx) {
+  MG_ASSERT(ctx->literal().size() == 2);
+  return std::pair{ctx->literal(0)->accept(this).as<Expression *>(), ctx->literal(1)->accept(this).as<Expression *>()};
+}
+
+antlrcpp::Any CypherMainVisitor::visitConfigMap(MemgraphCypher::ConfigMapContext *ctx) {
+  std::unordered_map<Expression *, Expression *> map;
+  for (auto *key_value_pair : ctx->configKeyValuePair()) {
+    // If the queries are cached, then only the stripped query is parsed, so the actual keys cannot be determined
+    // here. That means duplicates cannot be checked.
+    map.insert(key_value_pair->accept(this).as<std::pair<Expression *, Expression *>>());
+  }
+  return map;
+}
+
+antlrcpp::Any CypherMainVisitor::visitKafkaCreateStream(MemgraphCypher::KafkaCreateStreamContext *ctx) {
   auto *stream_query = storage_->Create<StreamQuery>();
   stream_query->action_ = StreamQuery::Action::CREATE_STREAM;
+  stream_query->type_ = StreamQuery::Type::KAFKA;
   stream_query->stream_name_ = ctx->streamName()->symbolicName()->accept(this).as<std::string>();
 
-  auto *topic_names_ctx = ctx->topicNames();
-  MG_ASSERT(topic_names_ctx != nullptr);
-  auto topic_names = topic_names_ctx->symbolicNameWithDotsAndMinus();
-  MG_ASSERT(!topic_names.empty());
-  stream_query->topic_names_.reserve(topic_names.size());
-  std::transform(topic_names.begin(), topic_names.end(), std::back_inserter(stream_query->topic_names_),
-                 [this](auto *topic_name) { return JoinSymbolicNamesWithDotsAndMinus(*this, *topic_name); });
+  for (auto *create_config_ctx : ctx->kafkaCreateStreamConfig()) {
+    create_config_ctx->accept(this);
+  }
 
-  stream_query->transform_name_ = JoinSymbolicNames(this, ctx->transformationName->symbolicName());
+  MapConfig<true, std::vector<std::string>, Expression *>(memory_, KafkaConfigKey::TOPICS, stream_query->topic_names_);
+  MapConfig<false, std::string>(memory_, KafkaConfigKey::CONSUMER_GROUP, stream_query->consumer_group_);
+  MapConfig<false, Expression *>(memory_, KafkaConfigKey::BOOTSTRAP_SERVERS, stream_query->bootstrap_servers_);
+  MapConfig<false, std::unordered_map<Expression *, Expression *>>(memory_, KafkaConfigKey::CONFIGS,
+                                                                   stream_query->configs_);
+  MapConfig<false, std::unordered_map<Expression *, Expression *>>(memory_, KafkaConfigKey::CREDENTIALS,
+                                                                   stream_query->credentials_);
+
+  MapCommonStreamConfigs(memory_, *stream_query);
+
+  return stream_query;
+}
+
+namespace {
+void ThrowIfExists(const auto &map, const EnumUint8 auto &enum_key) {
+  const auto key = static_cast<uint8_t>(enum_key);
+  if (map.contains(key)) {
+    throw SemanticException("{} defined multiple times in the query", ToString(enum_key));
+  }
+}
+
+void GetTopicNames(auto &destination, MemgraphCypher::TopicNamesContext *topic_names_ctx,
+                   antlr4::tree::ParseTreeVisitor &visitor) {
+  MG_ASSERT(topic_names_ctx != nullptr);
+  if (auto *symbolic_topic_names_ctx = topic_names_ctx->symbolicTopicNames()) {
+    destination = TopicNamesFromSymbols(visitor, symbolic_topic_names_ctx->symbolicNameWithDotsAndMinus());
+  } else {
+    if (!topic_names_ctx->literal()->StringLiteral()) {
+      throw SemanticException("Topic names should be defined as a string literal or as symbolic names");
+    }
+    destination = topic_names_ctx->accept(&visitor).as<Expression *>();
+  }
+}
+}  // namespace
+
+antlrcpp::Any CypherMainVisitor::visitKafkaCreateStreamConfig(MemgraphCypher::KafkaCreateStreamConfigContext *ctx) {
+  if (ctx->commonCreateStreamConfig()) {
+    return ctx->commonCreateStreamConfig()->accept(this);
+  }
+
+  if (ctx->TOPICS()) {
+    ThrowIfExists(memory_, KafkaConfigKey::TOPICS);
+    constexpr auto topics_key = static_cast<uint8_t>(KafkaConfigKey::TOPICS);
+    GetTopicNames(memory_[topics_key], ctx->topicNames(), *this);
+    return {};
+  }
 
   if (ctx->CONSUMER_GROUP()) {
-    stream_query->consumer_group_ = JoinSymbolicNamesWithDotsAndMinus(*this, *ctx->consumerGroup);
+    ThrowIfExists(memory_, KafkaConfigKey::CONSUMER_GROUP);
+    constexpr auto consumer_group_key = static_cast<uint8_t>(KafkaConfigKey::CONSUMER_GROUP);
+    memory_[consumer_group_key] = JoinSymbolicNamesWithDotsAndMinus(*this, *ctx->consumerGroup);
+    return {};
+  }
+
+  if (ctx->CONFIGS()) {
+    ThrowIfExists(memory_, KafkaConfigKey::CONFIGS);
+    constexpr auto configs_key = static_cast<uint8_t>(KafkaConfigKey::CONFIGS);
+    memory_.emplace(configs_key, ctx->configsMap->accept(this).as<std::unordered_map<Expression *, Expression *>>());
+    return {};
+  }
+
+  if (ctx->CREDENTIALS()) {
+    ThrowIfExists(memory_, KafkaConfigKey::CREDENTIALS);
+    constexpr auto credentials_key = static_cast<uint8_t>(KafkaConfigKey::CREDENTIALS);
+    memory_.emplace(credentials_key,
+                    ctx->credentialsMap->accept(this).as<std::unordered_map<Expression *, Expression *>>());
+    return {};
+  }
+
+  MG_ASSERT(ctx->BOOTSTRAP_SERVERS());
+  ThrowIfExists(memory_, KafkaConfigKey::BOOTSTRAP_SERVERS);
+  if (!ctx->bootstrapServers->StringLiteral()) {
+    throw SemanticException("Bootstrap servers should be a string!");
+  }
+
+  const auto bootstrap_servers_key = static_cast<uint8_t>(KafkaConfigKey::BOOTSTRAP_SERVERS);
+  memory_[bootstrap_servers_key] = ctx->bootstrapServers->accept(this).as<Expression *>();
+  return {};
+}
+
+namespace {
+GENERATE_STREAM_CONFIG_KEY_ENUM(Pulsar, TOPICS, SERVICE_URL);
+
+std::string_view ToString(const PulsarConfigKey key) {
+  switch (key) {
+    case PulsarConfigKey::TOPICS:
+      return "TOPICS";
+    case PulsarConfigKey::SERVICE_URL:
+      return "SERVICE_URL";
+  }
+}
+}  // namespace
+
+antlrcpp::Any CypherMainVisitor::visitPulsarCreateStream(MemgraphCypher::PulsarCreateStreamContext *ctx) {
+  auto *stream_query = storage_->Create<StreamQuery>();
+  stream_query->action_ = StreamQuery::Action::CREATE_STREAM;
+  stream_query->type_ = StreamQuery::Type::PULSAR;
+  stream_query->stream_name_ = ctx->streamName()->symbolicName()->accept(this).as<std::string>();
+
+  for (auto *create_config_ctx : ctx->pulsarCreateStreamConfig()) {
+    create_config_ctx->accept(this);
+  }
+
+  MapConfig<true, std::vector<std::string>, Expression *>(memory_, PulsarConfigKey::TOPICS, stream_query->topic_names_);
+  MapConfig<false, Expression *>(memory_, PulsarConfigKey::SERVICE_URL, stream_query->service_url_);
+
+  MapCommonStreamConfigs(memory_, *stream_query);
+
+  return stream_query;
+}
+
+antlrcpp::Any CypherMainVisitor::visitPulsarCreateStreamConfig(MemgraphCypher::PulsarCreateStreamConfigContext *ctx) {
+  if (ctx->commonCreateStreamConfig()) {
+    return ctx->commonCreateStreamConfig()->accept(this);
+  }
+
+  if (ctx->TOPICS()) {
+    ThrowIfExists(memory_, PulsarConfigKey::TOPICS);
+    const auto topics_key = static_cast<uint8_t>(PulsarConfigKey::TOPICS);
+    GetTopicNames(memory_[topics_key], ctx->topicNames(), *this);
+    return {};
+  }
+
+  MG_ASSERT(ctx->SERVICE_URL());
+  ThrowIfExists(memory_, PulsarConfigKey::SERVICE_URL);
+  if (!ctx->serviceUrl->StringLiteral()) {
+    throw SemanticException("Service URL must be a string!");
+  }
+  const auto service_url_key = static_cast<uint8_t>(PulsarConfigKey::SERVICE_URL);
+  memory_[service_url_key] = ctx->serviceUrl->accept(this).as<Expression *>();
+  return {};
+}
+
+antlrcpp::Any CypherMainVisitor::visitCommonCreateStreamConfig(MemgraphCypher::CommonCreateStreamConfigContext *ctx) {
+  if (ctx->TRANSFORM()) {
+    ThrowIfExists(memory_, CommonStreamConfigKey::TRANSFORM);
+    const auto transform_key = static_cast<uint8_t>(CommonStreamConfigKey::TRANSFORM);
+    memory_[transform_key] = JoinSymbolicNames(this, ctx->transformationName->symbolicName());
+    return {};
   }
 
   if (ctx->BATCH_INTERVAL()) {
+    ThrowIfExists(memory_, CommonStreamConfigKey::BATCH_INTERVAL);
     if (!ctx->batchInterval->numberLiteral() || !ctx->batchInterval->numberLiteral()->integerLiteral()) {
-      throw SemanticException("Batch interval should be an integer literal!");
+      throw SemanticException("Batch interval must be an integer literal!");
     }
-    stream_query->batch_interval_ = ctx->batchInterval->accept(this);
+    const auto batch_interval_key = static_cast<uint8_t>(CommonStreamConfigKey::BATCH_INTERVAL);
+    memory_[batch_interval_key] = ctx->batchInterval->accept(this).as<Expression *>();
+    return {};
   }
 
-  if (ctx->BATCH_SIZE()) {
-    if (!ctx->batchSize->numberLiteral() || !ctx->batchSize->numberLiteral()->integerLiteral()) {
-      throw SemanticException("Batch size should be an integer literal!");
-    }
-    stream_query->batch_size_ = ctx->batchSize->accept(this);
+  MG_ASSERT(ctx->BATCH_SIZE());
+  ThrowIfExists(memory_, CommonStreamConfigKey::BATCH_SIZE);
+  if (!ctx->batchSize->numberLiteral() || !ctx->batchSize->numberLiteral()->integerLiteral()) {
+    throw SemanticException("Batch size must be an integer literal!");
   }
-  if (ctx->BOOTSTRAP_SERVERS()) {
-    if (!ctx->bootstrapServers->StringLiteral()) {
-      throw SemanticException("Bootstrap servers should be a string!");
-    }
-    stream_query->bootstrap_servers_ = ctx->bootstrapServers->accept(this);
-  }
-
-  return stream_query;
+  const auto batch_size_key = static_cast<uint8_t>(CommonStreamConfigKey::BATCH_SIZE);
+  memory_[batch_size_key] = ctx->batchSize->accept(this).as<Expression *>();
+  return {};
 }
 
 antlrcpp::Any CypherMainVisitor::visitDropStream(MemgraphCypher::DropStreamContext *ctx) {

src/query/frontend/ast/cypher_main_visitor.hpp

@@ -1,4 +1,4 @@
-// Copyright 2021 Memgraph Ltd.
+// 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
@@ -269,6 +269,41 @@ class CypherMainVisitor : public antlropencypher::MemgraphCypherBaseVisitor {
    */
   antlrcpp::Any visitCreateStream(MemgraphCypher::CreateStreamContext *ctx) override;
 
+  /**
+   * @return StreamQuery*
+   */
+  antlrcpp::Any visitConfigKeyValuePair(MemgraphCypher::ConfigKeyValuePairContext *ctx) override;
+
+  /**
+   * @return StreamQuery*
+   */
+  antlrcpp::Any visitConfigMap(MemgraphCypher::ConfigMapContext *ctx) override;
+
+  /**
+   * @return StreamQuery*
+   */
+  antlrcpp::Any visitKafkaCreateStream(MemgraphCypher::KafkaCreateStreamContext *ctx) override;
+
+  /**
+   * @return StreamQuery*
+   */
+  antlrcpp::Any visitKafkaCreateStreamConfig(MemgraphCypher::KafkaCreateStreamConfigContext *ctx) override;
+
+  /**
+   * @return StreamQuery*
+   */
+  antlrcpp::Any visitPulsarCreateStreamConfig(MemgraphCypher::PulsarCreateStreamConfigContext *ctx) override;
+
+  /**
+   * @return StreamQuery*
+   */
+  antlrcpp::Any visitPulsarCreateStream(MemgraphCypher::PulsarCreateStreamContext *ctx) override;
+
+  /**
+   * @return StreamQuery*
+   */
+  antlrcpp::Any visitCommonCreateStreamConfig(MemgraphCypher::CommonCreateStreamConfigContext *ctx) override;
+
   /**
    * @return StreamQuery*
    */
@@ -824,6 +859,9 @@ class CypherMainVisitor : public antlropencypher::MemgraphCypherBaseVisitor {
   ParsingContext context_;
   AstStorage *storage_;
 
+  std::unordered_map<uint8_t, std::variant<Expression *, std::string, std::vector<std::string>,
+                                           std::unordered_map<Expression *, Expression *>>>
+      memory_;
   // Set of identifiers from queries.
   std::unordered_set<std::string> users_identifiers;
   // Identifiers that user didn't name.

src/query/frontend/opencypher/grammar/MemgraphCypher.g4

@@ -35,7 +35,9 @@ memgraphCypherKeyword : cypherKeyword
                       | COMMIT
                       | COMMITTED
                       | CONFIG
+                      | CONFIGS
                       | CONSUMER_GROUP
+                      | CREDENTIALS
                       | CSV
                       | DATA
                       | DELIMITER
@@ -295,15 +297,38 @@ symbolicNameWithMinus : symbolicName ( MINUS symbolicName )* ;
 
 symbolicNameWithDotsAndMinus: symbolicNameWithMinus ( DOT symbolicNameWithMinus )* ;
 
-topicNames : symbolicNameWithDotsAndMinus ( COMMA symbolicNameWithDotsAndMinus )* ;
+symbolicTopicNames : symbolicNameWithDotsAndMinus ( COMMA symbolicNameWithDotsAndMinus )* ;
 
-createStream : CREATE STREAM streamName
-               TOPICS topicNames
-               TRANSFORM transformationName=procedureName
-               ( CONSUMER_GROUP consumerGroup=symbolicNameWithDotsAndMinus ) ?
-               ( BATCH_INTERVAL batchInterval=literal ) ?
-               ( BATCH_SIZE batchSize=literal ) ?
-               ( BOOTSTRAP_SERVERS bootstrapServers=literal) ? ;
+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 ;
 

src/query/frontend/opencypher/grammar/MemgraphCypherLexer.g4

@@ -40,7 +40,9 @@ 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 ;
@@ -62,6 +64,7 @@ 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 ;
@@ -72,6 +75,7 @@ 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 ;
@@ -82,6 +86,7 @@ REVOKE              : R E V O K E ;
 ROLE                : R O L E ;
 ROLES               : R O L E S ;
 QUOTE               : Q U O T E ;
+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 ;

src/query/frontend/stripped_lexer_constants.hpp

@@ -199,7 +199,10 @@ const trie::Trie kKeywords = {"union",
                               "check",
                               "setting",
                               "settings",
-                              "bootstrap_servers"};
+                              "bootstrap_servers",
+                              "kafka",
+                              "pulsar",
+                              "service_url"};
 
 // Unicode codepoints that are allowed at the start of the unescaped name.
 const std::bitset<kBitsetSize> kUnescapedNameAllowedStarts(

src/query/interpreter.cpp

@@ -1,4 +1,4 @@
-// Copyright 2021 Memgraph Ltd.
+// 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
@@ -10,9 +10,14 @@
 // licenses/APL.txt.
 
 #include "query/interpreter.hpp"
+#include <fmt/core.h>
 
+#include <algorithm>
 #include <atomic>
 #include <chrono>
+#include <cstddef>
+#include <cstdint>
+#include <functional>
 #include <limits>
 #include <optional>
 
@@ -31,10 +36,11 @@
 #include "query/frontend/semantic/required_privileges.hpp"
 #include "query/frontend/semantic/symbol_generator.hpp"
 #include "query/interpret/eval.hpp"
+#include "query/metadata.hpp"
 #include "query/plan/planner.hpp"
 #include "query/plan/profile.hpp"
 #include "query/plan/vertex_count_cache.hpp"
-#include "query/streams.hpp"
+#include "query/stream/common.hpp"
 #include "query/trigger.hpp"
 #include "query/typed_value.hpp"
 #include "storage/v2/property_value.hpp"
@@ -52,6 +58,7 @@
 #include "utils/settings.hpp"
 #include "utils/string.hpp"
 #include "utils/tsc.hpp"
+#include "utils/variant_helpers.hpp"
 
 namespace EventCounter {
 extern Event ReadQuery;
@@ -86,7 +93,8 @@ void UpdateTypeCount(const plan::ReadWriteTypeChecker::RWType type) {
 
 struct Callback {
   std::vector<std::string> header;
-  std::function<std::vector<std::vector<TypedValue>>()> fn;
+  using CallbackFunction = std::function<std::vector<std::vector<TypedValue>>()>;
+  CallbackFunction fn;
   bool should_abort_query{false};
 };
 
@@ -405,7 +413,8 @@ Callback HandleAuthQuery(AuthQuery *auth_query, AuthQueryHandler *auth, const Pa
 }
 
 Callback HandleReplicationQuery(ReplicationQuery *repl_query, const Parameters &parameters,
-                                InterpreterContext *interpreter_context, DbAccessor *db_accessor) {
+                                InterpreterContext *interpreter_context, DbAccessor *db_accessor,
+                                std::vector<Notification> *notifications) {
   Frame frame(0);
   SymbolTable symbol_table;
   EvaluationContext evaluation_context;
@@ -423,11 +432,19 @@ Callback HandleReplicationQuery(ReplicationQuery *repl_query, const Parameters &
       if (port.IsInt()) {
         maybe_port = port.ValueInt();
       }
+      if (maybe_port == 7687 && repl_query->role_ == ReplicationQuery::ReplicationRole::REPLICA) {
+        notifications->emplace_back(SeverityLevel::WARNING, NotificationCode::REPLICA_PORT_WARNING,
+                                    "Be careful the replication port must be different from the memgraph port!");
+      }
       callback.fn = [handler = ReplQueryHandler{interpreter_context->db}, role = repl_query->role_,
                      maybe_port]() mutable {
         handler.SetReplicationRole(role, maybe_port);
         return std::vector<std::vector<TypedValue>>();
       };
+      notifications->emplace_back(
+          SeverityLevel::INFO, NotificationCode::SET_REPLICA,
+          fmt::format("Replica role set to {}.",
+                      repl_query->role_ == ReplicationQuery::ReplicationRole::MAIN ? "MAIN" : "REPLICA"));
       return callback;
     }
     case ReplicationQuery::Action::SHOW_REPLICATION_ROLE: {
@@ -461,6 +478,8 @@ Callback HandleReplicationQuery(ReplicationQuery *repl_query, const Parameters &
         handler.RegisterReplica(name, std::string(socket_address.ValueString()), sync_mode, maybe_timeout);
         return std::vector<std::vector<TypedValue>>();
       };
+      notifications->emplace_back(SeverityLevel::INFO, NotificationCode::REGISTER_REPLICA,
+                                  fmt::format("Replica {} is registered.", repl_query->replica_name_));
       return callback;
     }
     case ReplicationQuery::Action::DROP_REPLICA: {
@@ -469,6 +488,8 @@ Callback HandleReplicationQuery(ReplicationQuery *repl_query, const Parameters &
         handler.DropReplica(name);
         return std::vector<std::vector<TypedValue>>();
       };
+      notifications->emplace_back(SeverityLevel::INFO, NotificationCode::DROP_REPLICA,
+                                  fmt::format("Replica {} is dropped.", repl_query->replica_name_));
       return callback;
     }
     case ReplicationQuery::Action::SHOW_REPLICAS: {
@@ -511,9 +532,100 @@ std::optional<std::string> StringPointerToOptional(const std::string *str) {
   return str == nullptr ? std::nullopt : std::make_optional(*str);
 }
 
+stream::CommonStreamInfo GetCommonStreamInfo(StreamQuery *stream_query, ExpressionEvaluator &evaluator) {
+  return {
+      .batch_interval = GetOptionalValue<std::chrono::milliseconds>(stream_query->batch_interval_, evaluator)
+                            .value_or(stream::kDefaultBatchInterval),
+      .batch_size = GetOptionalValue<int64_t>(stream_query->batch_size_, evaluator).value_or(stream::kDefaultBatchSize),
+      .transformation_name = stream_query->transform_name_};
+}
+
+std::vector<std::string> EvaluateTopicNames(ExpressionEvaluator &evaluator,
+                                            std::variant<Expression *, std::vector<std::string>> topic_variant) {
+  return std::visit(utils::Overloaded{[&](Expression *expression) {
+                                        auto topic_names = expression->Accept(evaluator);
+                                        MG_ASSERT(topic_names.IsString());
+                                        return utils::Split(topic_names.ValueString(), ",");
+                                      },
+                                      [&](std::vector<std::string> topic_names) { return topic_names; }},
+                    std::move(topic_variant));
+}
+
+Callback::CallbackFunction GetKafkaCreateCallback(StreamQuery *stream_query, ExpressionEvaluator &evaluator,
+                                                  InterpreterContext *interpreter_context,
+                                                  const std::string *username) {
+  constexpr std::string_view kDefaultConsumerGroup = "mg_consumer";
+  std::string consumer_group{stream_query->consumer_group_.empty() ? kDefaultConsumerGroup
+                                                                   : stream_query->consumer_group_};
+
+  auto bootstrap = GetOptionalStringValue(stream_query->bootstrap_servers_, evaluator);
+  if (bootstrap && bootstrap->empty()) {
+    throw SemanticException("Bootstrap servers must not be an empty string!");
+  }
+  auto common_stream_info = GetCommonStreamInfo(stream_query, evaluator);
+
+  const auto get_config_map = [&evaluator](std::unordered_map<Expression *, Expression *> map,
+                                           std::string_view map_name) -> std::unordered_map<std::string, std::string> {
+    std::unordered_map<std::string, std::string> config_map;
+    for (const auto [key_expr, value_expr] : map) {
+      const auto key = key_expr->Accept(evaluator);
+      const auto value = value_expr->Accept(evaluator);
+      if (!key.IsString() || !value.IsString()) {
+        throw SemanticException("{} must contain only string keys and values!", map_name);
+      }
+      config_map.emplace(key.ValueString(), value.ValueString());
+    }
+    return config_map;
+  };
+
+  return [interpreter_context, stream_name = stream_query->stream_name_,
+          topic_names = EvaluateTopicNames(evaluator, stream_query->topic_names_),
+          consumer_group = std::move(consumer_group), common_stream_info = std::move(common_stream_info),
+          bootstrap_servers = std::move(bootstrap), owner = StringPointerToOptional(username),
+          configs = get_config_map(stream_query->configs_, "Configs"),
+          credentials = get_config_map(stream_query->credentials_, "Credentials")]() mutable {
+    std::string bootstrap = bootstrap_servers
+                                ? std::move(*bootstrap_servers)
+                                : std::string{interpreter_context->config.default_kafka_bootstrap_servers};
+    interpreter_context->streams.Create<query::stream::KafkaStream>(stream_name,
+                                                                    {.common_info = std::move(common_stream_info),
+                                                                     .topics = std::move(topic_names),
+                                                                     .consumer_group = std::move(consumer_group),
+                                                                     .bootstrap_servers = std::move(bootstrap),
+                                                                     .configs = std::move(configs),
+                                                                     .credentials = std::move(credentials)},
+                                                                    std::move(owner));
+
+    return std::vector<std::vector<TypedValue>>{};
+  };
+}
+
+Callback::CallbackFunction GetPulsarCreateCallback(StreamQuery *stream_query, ExpressionEvaluator &evaluator,
+                                                   InterpreterContext *interpreter_context,
+                                                   const std::string *username) {
+  auto service_url = GetOptionalStringValue(stream_query->service_url_, evaluator);
+  if (service_url && service_url->empty()) {
+    throw SemanticException("Service URL must not be an empty string!");
+  }
+  auto common_stream_info = GetCommonStreamInfo(stream_query, evaluator);
+  return [interpreter_context, stream_name = stream_query->stream_name_,
+          topic_names = EvaluateTopicNames(evaluator, stream_query->topic_names_),
+          common_stream_info = std::move(common_stream_info), service_url = std::move(service_url),
+          owner = StringPointerToOptional(username)]() mutable {
+    std::string url =
+        service_url ? std::move(*service_url) : std::string{interpreter_context->config.default_pulsar_service_url};
+    interpreter_context->streams.Create<query::stream::PulsarStream>(
+        stream_name,
+        {.common_info = std::move(common_stream_info), .topics = std::move(topic_names), .service_url = std::move(url)},
+        std::move(owner));
+
+    return std::vector<std::vector<TypedValue>>{};
+  };
+}
+
 Callback HandleStreamQuery(StreamQuery *stream_query, const Parameters &parameters,
                            InterpreterContext *interpreter_context, DbAccessor *db_accessor,
-                           const std::string *username) {
+                           const std::string *username, std::vector<Notification> *notifications) {
   Frame frame(0);
   SymbolTable symbol_table;
   EvaluationContext evaluation_context;
@@ -527,32 +639,16 @@ Callback HandleStreamQuery(StreamQuery *stream_query, const Parameters &paramete
   switch (stream_query->action_) {
     case StreamQuery::Action::CREATE_STREAM: {
       EventCounter::IncrementCounter(EventCounter::StreamsCreated);
-      constexpr std::string_view kDefaultConsumerGroup = "mg_consumer";
-      std::string consumer_group{stream_query->consumer_group_.empty() ? kDefaultConsumerGroup
-                                                                       : stream_query->consumer_group_};
-
-      auto bootstrap = GetOptionalStringValue(stream_query->bootstrap_servers_, evaluator);
-      if (bootstrap && bootstrap->empty()) {
-        throw SemanticException("Bootstrap servers must not be an empty string!");
+      switch (stream_query->type_) {
+        case StreamQuery::Type::KAFKA:
+          callback.fn = GetKafkaCreateCallback(stream_query, evaluator, interpreter_context, username);
+          break;
+        case StreamQuery::Type::PULSAR:
+          callback.fn = GetPulsarCreateCallback(stream_query, evaluator, interpreter_context, username);
+          break;
       }
-      callback.fn = [interpreter_context, stream_name = stream_query->stream_name_,
-                     topic_names = stream_query->topic_names_, consumer_group = std::move(consumer_group),
-                     batch_interval =
-                         GetOptionalValue<std::chrono::milliseconds>(stream_query->batch_interval_, evaluator),
-                     batch_size = GetOptionalValue<int64_t>(stream_query->batch_size_, evaluator),
-                     transformation_name = stream_query->transform_name_, bootstrap_servers = std::move(bootstrap),
-                     owner = StringPointerToOptional(username)]() mutable {
-        std::string bootstrap = bootstrap_servers ? std::move(*bootstrap_servers) : "";
-        interpreter_context->streams.Create(stream_name,
-                                            query::StreamInfo{.topics = std::move(topic_names),
-                                                              .consumer_group = std::move(consumer_group),
-                                                              .batch_interval = batch_interval,
-                                                              .batch_size = batch_size,
-                                                              .transformation_name = std::move(transformation_name),
-                                                              .owner = std::move(owner),
-                                                              .bootstrap_servers = std::move(bootstrap)});
-        return std::vector<std::vector<TypedValue>>{};
-      };
+      notifications->emplace_back(SeverityLevel::INFO, NotificationCode::CREATE_STREAM,
+                                  fmt::format("Created stream {}.", stream_query->stream_name_));
       return callback;
     }
     case StreamQuery::Action::START_STREAM: {
@@ -560,6 +656,8 @@ Callback HandleStreamQuery(StreamQuery *stream_query, const Parameters &paramete
         interpreter_context->streams.Start(stream_name);
         return std::vector<std::vector<TypedValue>>{};
       };
+      notifications->emplace_back(SeverityLevel::INFO, NotificationCode::START_STREAM,
+                                  fmt::format("Started stream {}.", stream_query->stream_name_));
       return callback;
     }
     case StreamQuery::Action::START_ALL_STREAMS: {
@@ -567,6 +665,7 @@ Callback HandleStreamQuery(StreamQuery *stream_query, const Parameters &paramete
         interpreter_context->streams.StartAll();
         return std::vector<std::vector<TypedValue>>{};
       };
+      notifications->emplace_back(SeverityLevel::INFO, NotificationCode::START_ALL_STREAMS, "Started all streams.");
       return callback;
     }
     case StreamQuery::Action::STOP_STREAM: {
@@ -574,6 +673,8 @@ Callback HandleStreamQuery(StreamQuery *stream_query, const Parameters &paramete
         interpreter_context->streams.Stop(stream_name);
         return std::vector<std::vector<TypedValue>>{};
       };
+      notifications->emplace_back(SeverityLevel::INFO, NotificationCode::STOP_STREAM,
+                                  fmt::format("Stopped stream {}.", stream_query->stream_name_));
       return callback;
     }
     case StreamQuery::Action::STOP_ALL_STREAMS: {
@@ -581,6 +682,7 @@ Callback HandleStreamQuery(StreamQuery *stream_query, const Parameters &paramete
         interpreter_context->streams.StopAll();
         return std::vector<std::vector<TypedValue>>{};
       };
+      notifications->emplace_back(SeverityLevel::INFO, NotificationCode::STOP_ALL_STREAMS, "Stopped all streams.");
       return callback;
     }
     case StreamQuery::Action::DROP_STREAM: {
@@ -588,59 +690,33 @@ Callback HandleStreamQuery(StreamQuery *stream_query, const Parameters &paramete
         interpreter_context->streams.Drop(stream_name);
         return std::vector<std::vector<TypedValue>>{};
       };
+      notifications->emplace_back(SeverityLevel::INFO, NotificationCode::DROP_STREAM,
+                                  fmt::format("Dropped stream {}.", stream_query->stream_name_));
       return callback;
     }
     case StreamQuery::Action::SHOW_STREAMS: {
-      callback.header = {"name",           "topics",
-                         "consumer_group", "batch_interval",
-                         "batch_size",     "transformation_name",
-                         "owner",          "bootstrap_servers",
-                         "is running"};
+      callback.header = {"name", "type", "batch_interval", "batch_size", "transformation_name", "owner", "is running"};
       callback.fn = [interpreter_context]() {
         auto streams_status = interpreter_context->streams.GetStreamInfo();
         std::vector<std::vector<TypedValue>> results;
         results.reserve(streams_status.size());
-        auto topics_as_typed_topics = [](const auto &topics) {
-          std::vector<TypedValue> typed_topics;
-          typed_topics.reserve(topics.size());
-          for (const auto &elem : topics) {
-            typed_topics.emplace_back(elem);
-          }
-          return typed_topics;
-        };
-
-        auto stream_info_as_typed_stream_info_emplace_in = [topics_as_typed_topics, interpreter_context](
-                                                               auto &typed_status, const auto &stream_info) {
-          typed_status.emplace_back(topics_as_typed_topics(stream_info.topics));
-          typed_status.emplace_back(stream_info.consumer_group);
-          if (stream_info.batch_interval.has_value()) {
-            typed_status.emplace_back(stream_info.batch_interval->count());
-          } else {
-            typed_status.emplace_back();
-          }
-          if (stream_info.batch_size.has_value()) {
-            typed_status.emplace_back(*stream_info.batch_size);
-          } else {
-            typed_status.emplace_back();
-          }
+        auto stream_info_as_typed_stream_info_emplace_in = [](auto &typed_status, const auto &stream_info) {
+          typed_status.emplace_back(stream_info.batch_interval.count());
+          typed_status.emplace_back(stream_info.batch_size);
           typed_status.emplace_back(stream_info.transformation_name);
-          if (stream_info.owner.has_value()) {
-            typed_status.emplace_back(*stream_info.owner);
-          } else {
-            typed_status.emplace_back();
-          }
-          if (stream_info.bootstrap_servers.empty()) {
-            typed_status.emplace_back(interpreter_context->streams.BootstrapServers());
-          } else {
-            typed_status.emplace_back(stream_info.bootstrap_servers);
-          }
         };
 
         for (const auto &status : streams_status) {
           std::vector<TypedValue> typed_status;
-          typed_status.reserve(8);
+          typed_status.reserve(7);
           typed_status.emplace_back(status.name);
+          typed_status.emplace_back(StreamSourceTypeToString(status.type));
           stream_info_as_typed_stream_info_emplace_in(typed_status, status.info);
+          if (status.owner.has_value()) {
+            typed_status.emplace_back(*status.owner);
+          } else {
+            typed_status.emplace_back();
+          }
           typed_status.emplace_back(status.is_running);
           results.push_back(std::move(typed_status));
         }
@@ -656,6 +732,8 @@ Callback HandleStreamQuery(StreamQuery *stream_query, const Parameters &paramete
                      batch_limit = GetOptionalValue<int64_t>(stream_query->batch_limit_, evaluator)]() mutable {
         return interpreter_context->streams.Check(stream_name, timeout, batch_limit);
       };
+      notifications->emplace_back(SeverityLevel::INFO, NotificationCode::CHECK_STREAM,
+                                  fmt::format("Checked stream {}.", stream_query->stream_name_));
       return callback;
     }
   }
@@ -830,14 +908,14 @@ std::optional<plan::ProfilingStatsWithTotalTime> PullPlan::Pull(AnyStream *strea
   // Also, we want to throw only when the query engine requests more memory and not the storage
   // so we add the exception to the allocator.
   // TODO (mferencevic): Tune the parameters accordingly.
-  utils::FixedSizePoolResource pool_memory(128, 1024, &monotonic_memory);
+  utils::PoolResource pool_memory(128, 1024, &monotonic_memory);
   std::optional<utils::LimitedMemoryResource> maybe_limited_resource;
 
   if (memory_limit_) {
-    maybe_limited_resource.emplace(utils::NewDeleteResource(), *memory_limit_);
+    maybe_limited_resource.emplace(&pool_memory, *memory_limit_);
     ctx_.evaluation_context.memory = &*maybe_limited_resource;
   } else {
-    ctx_.evaluation_context.memory = utils::NewDeleteResource();
+    ctx_.evaluation_context.memory = &pool_memory;
   }
 
   // Returns true if a result was pulled.
@@ -886,8 +964,19 @@ std::optional<plan::ProfilingStatsWithTotalTime> PullPlan::Pull(AnyStream *strea
   if (has_unsent_results_) {
     return std::nullopt;
   }
-
   summary->insert_or_assign("plan_execution_time", execution_time_.count());
+  // We are finished with pulling all the data, therefore we can send any
+  // metadata about the results i.e. notifications and statistics
+  const bool is_any_counter_set =
+      std::any_of(ctx_.execution_stats.counters.begin(), ctx_.execution_stats.counters.end(),
+                  [](const auto &counter) { return counter > 0; });
+  if (is_any_counter_set) {
+    std::map<std::string, TypedValue> stats;
+    for (size_t i = 0; i < ctx_.execution_stats.counters.size(); ++i) {
+      stats.emplace(ExecutionStatsKeyToString(ExecutionStats::Key(i)), ctx_.execution_stats.counters[i]);
+    }
+    summary->insert_or_assign("stats", std::move(stats));
+  }
   cursor_->Shutdown();
   ctx_.profile_execution_time = execution_time_;
   return GetStatsWithTotalTime(ctx_);
@@ -897,11 +986,8 @@ using RWType = plan::ReadWriteTypeChecker::RWType;
 }  // namespace
 
 InterpreterContext::InterpreterContext(storage::Storage *db, const InterpreterConfig config,
-                                       const std::filesystem::path &data_directory, std::string kafka_bootstrap_servers)
-    : db(db),
-      trigger_store(data_directory / "triggers"),
-      config(config),
-      streams{this, std::move(kafka_bootstrap_servers), data_directory / "streams"} {}
+                                       const std::filesystem::path &data_directory)
+    : db(db), trigger_store(data_directory / "triggers"), config(config), streams{this, data_directory / "streams"} {}
 
 Interpreter::Interpreter(InterpreterContext *interpreter_context) : interpreter_context_(interpreter_context) {
   MG_ASSERT(interpreter_context_, "Interpreter context must not be NULL");
@@ -971,7 +1057,7 @@ PreparedQuery Interpreter::PrepareTransactionQuery(std::string_view query_upper)
 
 PreparedQuery PrepareCypherQuery(ParsedQuery parsed_query, std::map<std::string, TypedValue> *summary,
                                  InterpreterContext *interpreter_context, DbAccessor *dba,
-                                 utils::MemoryResource *execution_memory,
+                                 utils::MemoryResource *execution_memory, std::vector<Notification> *notifications,
                                  TriggerContextCollector *trigger_context_collector = nullptr) {
   auto *cypher_query = utils::Downcast<CypherQuery>(parsed_query.query);
 
@@ -986,6 +1072,15 @@ PreparedQuery PrepareCypherQuery(ParsedQuery parsed_query, std::map<std::string,
     spdlog::info("Running query with memory limit of {}", utils::GetReadableSize(*memory_limit));
   }
 
+  if (const auto &clauses = cypher_query->single_query_->clauses_; std::any_of(
+          clauses.begin(), clauses.end(), [](const auto *clause) { return clause->GetTypeInfo() == LoadCsv::kType; })) {
+    notifications->emplace_back(
+        SeverityLevel::INFO, NotificationCode::LOAD_CSV_TIP,
+        "It's important to note that the parser parses the values as strings. It's up to the user to "
+        "convert the parsed row values to the appropriate type. This can be done using the built-in "
+        "conversion functions such as ToInteger, ToFloat, ToBoolean etc.");
+  }
+
   auto plan = CypherQueryToPlan(parsed_query.stripped_query.hash(), std::move(parsed_query.ast_storage), cypher_query,
                                 parsed_query.parameters,
                                 parsed_query.is_cacheable ? &interpreter_context->plan_cache : nullptr, dba);
@@ -1006,7 +1101,6 @@ PreparedQuery PrepareCypherQuery(ParsedQuery parsed_query, std::map<std::string,
     header.push_back(
         utils::FindOr(parsed_query.stripped_query.named_expressions(), symbol.token_position(), symbol.name()).first);
   }
-
   auto pull_plan = std::make_shared<PullPlan>(plan, parsed_query.parameters, false, dba, interpreter_context,
                                               execution_memory, trigger_context_collector, memory_limit);
   return PreparedQuery{std::move(header), std::move(parsed_query.required_privileges),
@@ -1164,14 +1258,13 @@ PreparedQuery PrepareDumpQuery(ParsedQuery parsed_query, std::map<std::string, T
 }
 
 PreparedQuery PrepareIndexQuery(ParsedQuery parsed_query, bool in_explicit_transaction,
-                                std::map<std::string, TypedValue> *summary, InterpreterContext *interpreter_context,
-                                utils::MemoryResource *execution_memory) {
+                                std::vector<Notification> *notifications, InterpreterContext *interpreter_context) {
   if (in_explicit_transaction) {
     throw IndexInMulticommandTxException();
   }
 
   auto *index_query = utils::Downcast<IndexQuery>(parsed_query.query);
-  std::function<void()> handler;
+  std::function<void(Notification &)> handler;
 
   // Creating an index influences computed plan costs.
   auto invalidate_plan_cache = [plan_cache = &interpreter_context->plan_cache] {
@@ -1182,26 +1275,45 @@ PreparedQuery PrepareIndexQuery(ParsedQuery parsed_query, bool in_explicit_trans
   };
 
   auto label = interpreter_context->db->NameToLabel(index_query->label_.name);
+
   std::vector<storage::PropertyId> properties;
+  std::vector<std::string> properties_string;
   properties.reserve(index_query->properties_.size());
+  properties_string.reserve(index_query->properties_.size());
   for (const auto &prop : index_query->properties_) {
     properties.push_back(interpreter_context->db->NameToProperty(prop.name));
+    properties_string.push_back(prop.name);
   }
+  auto properties_stringified = utils::Join(properties_string, ", ");
 
   if (properties.size() > 1) {
     throw utils::NotYetImplemented("index on multiple properties");
   }
 
+  Notification index_notification(SeverityLevel::INFO);
   switch (index_query->action_) {
     case IndexQuery::Action::CREATE: {
-      handler = [interpreter_context, label, properties = std::move(properties),
-                 invalidate_plan_cache = std::move(invalidate_plan_cache)] {
+      index_notification.code = NotificationCode::CREATE_INDEX;
+      index_notification.title =
+          fmt::format("Created index on label {} on properties {}.", index_query->label_.name, properties_stringified);
+
+      handler = [interpreter_context, label, properties_stringified = std::move(properties_stringified),
+                 label_name = index_query->label_.name, properties = std::move(properties),
+                 invalidate_plan_cache = std::move(invalidate_plan_cache)](Notification &index_notification) {
         if (properties.empty()) {
-          interpreter_context->db->CreateIndex(label);
+          if (!interpreter_context->db->CreateIndex(label)) {
+            index_notification.code = NotificationCode::EXISTANT_INDEX;
+            index_notification.title =
+                fmt::format("Index on label {} on properties {} already exists.", label_name, properties_stringified);
+          }
           EventCounter::IncrementCounter(EventCounter::LabelIndexCreated);
         } else {
           MG_ASSERT(properties.size() == 1U);
-          interpreter_context->db->CreateIndex(label, properties[0]);
+          if (!interpreter_context->db->CreateIndex(label, properties[0])) {
+            index_notification.code = NotificationCode::EXISTANT_INDEX;
+            index_notification.title =
+                fmt::format("Index on label {} on properties {} already exists.", label_name, properties_stringified);
+          }
           EventCounter::IncrementCounter(EventCounter::LabelPropertyIndexCreated);
         }
         invalidate_plan_cache();
@@ -1209,13 +1321,25 @@ PreparedQuery PrepareIndexQuery(ParsedQuery parsed_query, bool in_explicit_trans
       break;
     }
     case IndexQuery::Action::DROP: {
-      handler = [interpreter_context, label, properties = std::move(properties),
-                 invalidate_plan_cache = std::move(invalidate_plan_cache)] {
+      index_notification.code = NotificationCode::DROP_INDEX;
+      index_notification.title = fmt::format("Dropped index on label {} on properties {}.", index_query->label_.name,
+                                             utils::Join(properties_string, ", "));
+      handler = [interpreter_context, label, properties_stringified = std::move(properties_stringified),
+                 label_name = index_query->label_.name, properties = std::move(properties),
+                 invalidate_plan_cache = std::move(invalidate_plan_cache)](Notification &index_notification) {
         if (properties.empty()) {
-          interpreter_context->db->DropIndex(label);
+          if (!interpreter_context->db->DropIndex(label)) {
+            index_notification.code = NotificationCode::NONEXISTANT_INDEX;
+            index_notification.title =
+                fmt::format("Index on label {} on properties {} doesn't exist.", label_name, properties_stringified);
+          }
         } else {
           MG_ASSERT(properties.size() == 1U);
-          interpreter_context->db->DropIndex(label, properties[0]);
+          if (!interpreter_context->db->DropIndex(label, properties[0])) {
+            index_notification.code = NotificationCode::NONEXISTANT_INDEX;
+            index_notification.title =
+                fmt::format("Index on label {} on properties {} doesn't exist.", label_name, properties_stringified);
+          }
         }
         invalidate_plan_cache();
       };
@@ -1223,13 +1347,16 @@ PreparedQuery PrepareIndexQuery(ParsedQuery parsed_query, bool in_explicit_trans
     }
   }
 
-  return PreparedQuery{{},
-                       std::move(parsed_query.required_privileges),
-                       [handler = std::move(handler)](AnyStream *stream, std::optional<int>) {
-                         handler();
-                         return QueryHandlerResult::NOTHING;
-                       },
-                       RWType::W};
+  return PreparedQuery{
+      {},
+      std::move(parsed_query.required_privileges),
+      [handler = std::move(handler), notifications, index_notification = std::move(index_notification)](
+          AnyStream * /*stream*/, std::optional<int> /*unused*/) mutable {
+        handler(index_notification);
+        notifications->push_back(index_notification);
+        return QueryHandlerResult::NOTHING;
+      },
+      RWType::W};
 }
 
 PreparedQuery PrepareAuthQuery(ParsedQuery parsed_query, bool in_explicit_transaction,
@@ -1269,13 +1396,15 @@ PreparedQuery PrepareAuthQuery(ParsedQuery parsed_query, bool in_explicit_transa
 }
 
 PreparedQuery PrepareReplicationQuery(ParsedQuery parsed_query, const bool in_explicit_transaction,
-                                      InterpreterContext *interpreter_context, DbAccessor *dba) {
+                                      std::vector<Notification> *notifications, InterpreterContext *interpreter_context,
+                                      DbAccessor *dba) {
   if (in_explicit_transaction) {
     throw ReplicationModificationInMulticommandTxException();
   }
 
   auto *replication_query = utils::Downcast<ReplicationQuery>(parsed_query.query);
-  auto callback = HandleReplicationQuery(replication_query, parsed_query.parameters, interpreter_context, dba);
+  auto callback =
+      HandleReplicationQuery(replication_query, parsed_query.parameters, interpreter_context, dba, notifications);
 
   return PreparedQuery{callback.header, std::move(parsed_query.required_privileges),
                        [callback_fn = std::move(callback.fn), pull_plan = std::shared_ptr<PullPlanVector>{nullptr}](
@@ -1424,8 +1553,8 @@ Callback ShowTriggers(InterpreterContext *interpreter_context) {
 }
 
 PreparedQuery PrepareTriggerQuery(ParsedQuery parsed_query, const bool in_explicit_transaction,
-                                  InterpreterContext *interpreter_context, DbAccessor *dba,
-                                  const std::map<std::string, storage::PropertyValue> &user_parameters,
+                                  std::vector<Notification> *notifications, InterpreterContext *interpreter_context,
+                                  DbAccessor *dba, const std::map<std::string, storage::PropertyValue> &user_parameters,
                                   const std::string *username) {
   if (in_explicit_transaction) {
     throw TriggerModificationInMulticommandTxException();
@@ -1434,27 +1563,36 @@ PreparedQuery PrepareTriggerQuery(ParsedQuery parsed_query, const bool in_explic
   auto *trigger_query = utils::Downcast<TriggerQuery>(parsed_query.query);
   MG_ASSERT(trigger_query);
 
-  auto callback = [trigger_query, interpreter_context, dba, &user_parameters,
-                   owner = StringPointerToOptional(username)]() mutable {
+  std::optional<Notification> trigger_notification;
+  auto callback = std::invoke([trigger_query, interpreter_context, dba, &user_parameters,
+                               owner = StringPointerToOptional(username), &trigger_notification]() mutable {
     switch (trigger_query->action_) {
       case TriggerQuery::Action::CREATE_TRIGGER:
+        trigger_notification.emplace(SeverityLevel::INFO, NotificationCode::CREATE_TRIGGER,
+                                     fmt::format("Created trigger {}.", trigger_query->trigger_name_));
         EventCounter::IncrementCounter(EventCounter::TriggersCreated);
         return CreateTrigger(trigger_query, user_parameters, interpreter_context, dba, std::move(owner));
       case TriggerQuery::Action::DROP_TRIGGER:
+        trigger_notification.emplace(SeverityLevel::INFO, NotificationCode::DROP_TRIGGER,
+                                     fmt::format("Dropped trigger {}.", trigger_query->trigger_name_));
         return DropTrigger(trigger_query, interpreter_context);
       case TriggerQuery::Action::SHOW_TRIGGERS:
         return ShowTriggers(interpreter_context);
     }
-  }();
+  });
 
   return PreparedQuery{std::move(callback.header), std::move(parsed_query.required_privileges),
-                       [callback_fn = std::move(callback.fn), pull_plan = std::shared_ptr<PullPlanVector>{nullptr}](
+                       [callback_fn = std::move(callback.fn), pull_plan = std::shared_ptr<PullPlanVector>{nullptr},
+                        trigger_notification = std::move(trigger_notification), notifications](
                            AnyStream *stream, std::optional<int> n) mutable -> std::optional<QueryHandlerResult> {
                          if (UNLIKELY(!pull_plan)) {
                            pull_plan = std::make_shared<PullPlanVector>(callback_fn());
                          }
 
                          if (pull_plan->Pull(stream, n)) {
+                           if (trigger_notification) {
+                             notifications->push_back(std::move(*trigger_notification));
+                           }
                            return QueryHandlerResult::COMMIT;
                          }
                          return std::nullopt;
@@ -1465,8 +1603,9 @@ PreparedQuery PrepareTriggerQuery(ParsedQuery parsed_query, const bool in_explic
 }
 
 PreparedQuery PrepareStreamQuery(ParsedQuery parsed_query, const bool in_explicit_transaction,
-                                 InterpreterContext *interpreter_context, DbAccessor *dba,
-                                 const std::map<std::string, storage::PropertyValue> &user_parameters,
+                                 std::vector<Notification> *notifications, InterpreterContext *interpreter_context,
+                                 DbAccessor *dba,
+                                 const std::map<std::string, storage::PropertyValue> & /*user_parameters*/,
                                  const std::string *username) {
   if (in_explicit_transaction) {
     throw StreamQueryInMulticommandTxException();
@@ -1474,7 +1613,8 @@ PreparedQuery PrepareStreamQuery(ParsedQuery parsed_query, const bool in_explici
 
   auto *stream_query = utils::Downcast<StreamQuery>(parsed_query.query);
   MG_ASSERT(stream_query);
-  auto callback = HandleStreamQuery(stream_query, parsed_query.parameters, interpreter_context, dba, username);
+  auto callback =
+      HandleStreamQuery(stream_query, parsed_query.parameters, interpreter_context, dba, username, notifications);
 
   return PreparedQuery{std::move(callback.header), std::move(parsed_query.required_privileges),
                        [callback_fn = std::move(callback.fn), pull_plan = std::shared_ptr<PullPlanVector>{nullptr}](
@@ -1674,24 +1814,31 @@ PreparedQuery PrepareInfoQuery(ParsedQuery parsed_query, bool in_explicit_transa
 }
 
 PreparedQuery PrepareConstraintQuery(ParsedQuery parsed_query, bool in_explicit_transaction,
-                                     std::map<std::string, TypedValue> *summary,
-                                     InterpreterContext *interpreter_context, utils::MemoryResource *execution_memory) {
+                                     std::vector<Notification> *notifications,
+                                     InterpreterContext *interpreter_context) {
   if (in_explicit_transaction) {
     throw ConstraintInMulticommandTxException();
   }
 
   auto *constraint_query = utils::Downcast<ConstraintQuery>(parsed_query.query);
-  std::function<void()> handler;
+  std::function<void(Notification &)> handler;
 
   auto label = interpreter_context->db->NameToLabel(constraint_query->constraint_.label.name);
   std::vector<storage::PropertyId> properties;
+  std::vector<std::string> properties_string;
   properties.reserve(constraint_query->constraint_.properties.size());
+  properties_string.reserve(constraint_query->constraint_.properties.size());
   for (const auto &prop : constraint_query->constraint_.properties) {
     properties.push_back(interpreter_context->db->NameToProperty(prop.name));
+    properties_string.push_back(prop.name);
   }
+  auto properties_stringified = utils::Join(properties_string, ", ");
 
+  Notification constraint_notification(SeverityLevel::INFO);
   switch (constraint_query->action_type_) {
     case ConstraintQuery::ActionType::CREATE: {
+      constraint_notification.code = NotificationCode::CREATE_CONSTRAINT;
+
       switch (constraint_query->constraint_.type) {
         case Constraint::Type::NODE_KEY:
           throw utils::NotYetImplemented("Node key constraints");
@@ -1699,7 +1846,11 @@ PreparedQuery PrepareConstraintQuery(ParsedQuery parsed_query, bool in_explicit_
           if (properties.empty() || properties.size() > 1) {
             throw SyntaxException("Exactly one property must be used for existence constraints.");
           }
-          handler = [interpreter_context, label, properties = std::move(properties)] {
+          constraint_notification.title = fmt::format("Created EXISTS constraint on label {} on properties {}.",
+                                                      constraint_query->constraint_.label.name, properties_stringified);
+          handler = [interpreter_context, label, label_name = constraint_query->constraint_.label.name,
+                     properties_stringified = std::move(properties_stringified),
+                     properties = std::move(properties)](Notification &constraint_notification) {
             auto res = interpreter_context->db->CreateExistenceConstraint(label, properties[0]);
             if (res.HasError()) {
               auto violation = res.GetError();
@@ -1711,6 +1862,11 @@ PreparedQuery PrepareConstraintQuery(ParsedQuery parsed_query, bool in_explicit_
                   "existing node violates it.",
                   label_name, property_name);
             }
+            if (res.HasValue() && !res.GetValue()) {
+              constraint_notification.code = NotificationCode::EXISTANT_CONSTRAINT;
+              constraint_notification.title = fmt::format(
+                  "Constraint EXISTS on label {} on properties {} already exists.", label_name, properties_stringified);
+            }
           };
           break;
         case Constraint::Type::UNIQUE:
@@ -1721,7 +1877,12 @@ PreparedQuery PrepareConstraintQuery(ParsedQuery parsed_query, bool in_explicit_
           if (property_set.size() != properties.size()) {
             throw SyntaxException("The given set of properties contains duplicates.");
           }
-          handler = [interpreter_context, label, property_set = std::move(property_set)] {
+          constraint_notification.title =
+              fmt::format("Created UNIQUE constraint on label {} on properties {}.",
+                          constraint_query->constraint_.label.name, utils::Join(properties_string, ", "));
+          handler = [interpreter_context, label, label_name = constraint_query->constraint_.label.name,
+                     properties_stringified = std::move(properties_stringified),
+                     property_set = std::move(property_set)](Notification &constraint_notification) {
             auto res = interpreter_context->db->CreateUniqueConstraint(label, property_set);
             if (res.HasError()) {
               auto violation = res.GetError();
@@ -1735,29 +1896,33 @@ PreparedQuery PrepareConstraintQuery(ParsedQuery parsed_query, bool in_explicit_
                   "Unable to create unique constraint :{}({}), because an "
                   "existing node violates it.",
                   label_name, property_names_stream.str());
-            } else {
-              switch (res.GetValue()) {
-                case storage::UniqueConstraints::CreationStatus::EMPTY_PROPERTIES:
-                  throw SyntaxException(
-                      "At least one property must be used for unique "
-                      "constraints.");
-                  break;
-                case storage::UniqueConstraints::CreationStatus::PROPERTIES_SIZE_LIMIT_EXCEEDED:
-                  throw SyntaxException(
-                      "Too many properties specified. Limit of {} properties "
-                      "for unique constraints is exceeded.",
-                      storage::kUniqueConstraintsMaxProperties);
-                  break;
-                case storage::UniqueConstraints::CreationStatus::ALREADY_EXISTS:
-                case storage::UniqueConstraints::CreationStatus::SUCCESS:
-                  break;
-              }
+            }
+            switch (res.GetValue()) {
+              case storage::UniqueConstraints::CreationStatus::EMPTY_PROPERTIES:
+                throw SyntaxException(
+                    "At least one property must be used for unique "
+                    "constraints.");
+              case storage::UniqueConstraints::CreationStatus::PROPERTIES_SIZE_LIMIT_EXCEEDED:
+                throw SyntaxException(
+                    "Too many properties specified. Limit of {} properties "
+                    "for unique constraints is exceeded.",
+                    storage::kUniqueConstraintsMaxProperties);
+              case storage::UniqueConstraints::CreationStatus::ALREADY_EXISTS:
+                constraint_notification.code = NotificationCode::EXISTANT_CONSTRAINT;
+                constraint_notification.title =
+                    fmt::format("Constraint UNIQUE on label {} on properties {} already exists.", label_name,
+                                properties_stringified);
+                break;
+              case storage::UniqueConstraints::CreationStatus::SUCCESS:
+                break;
             }
           };
           break;
       }
     } break;
     case ConstraintQuery::ActionType::DROP: {
+      constraint_notification.code = NotificationCode::DROP_CONSTRAINT;
+
       switch (constraint_query->constraint_.type) {
         case Constraint::Type::NODE_KEY:
           throw utils::NotYetImplemented("Node key constraints");
@@ -1765,8 +1930,17 @@ PreparedQuery PrepareConstraintQuery(ParsedQuery parsed_query, bool in_explicit_
           if (properties.empty() || properties.size() > 1) {
             throw SyntaxException("Exactly one property must be used for existence constraints.");
           }
-          handler = [interpreter_context, label, properties = std::move(properties)] {
-            interpreter_context->db->DropExistenceConstraint(label, properties[0]);
+          constraint_notification.title =
+              fmt::format("Dropped EXISTS constraint on label {} on properties {}.",
+                          constraint_query->constraint_.label.name, utils::Join(properties_string, ", "));
+          handler = [interpreter_context, label, label_name = constraint_query->constraint_.label.name,
+                     properties_stringified = std::move(properties_stringified),
+                     properties = std::move(properties)](Notification &constraint_notification) {
+            if (!interpreter_context->db->DropExistenceConstraint(label, properties[0])) {
+              constraint_notification.code = NotificationCode::NONEXISTANT_CONSTRAINT;
+              constraint_notification.title = fmt::format(
+                  "Constraint EXISTS on label {} on properties {} doesn't exist.", label_name, properties_stringified);
+            }
             return std::vector<std::vector<TypedValue>>();
           };
           break;
@@ -1778,7 +1952,12 @@ PreparedQuery PrepareConstraintQuery(ParsedQuery parsed_query, bool in_explicit_
           if (property_set.size() != properties.size()) {
             throw SyntaxException("The given set of properties contains duplicates.");
           }
-          handler = [interpreter_context, label, property_set = std::move(property_set)] {
+          constraint_notification.title =
+              fmt::format("Dropped UNIQUE constraint on label {} on properties {}.",
+                          constraint_query->constraint_.label.name, utils::Join(properties_string, ", "));
+          handler = [interpreter_context, label, label_name = constraint_query->constraint_.label.name,
+                     properties_stringified = std::move(properties_stringified),
+                     property_set = std::move(property_set)](Notification &constraint_notification) {
             auto res = interpreter_context->db->DropUniqueConstraint(label, property_set);
             switch (res) {
               case storage::UniqueConstraints::DeletionStatus::EMPTY_PROPERTIES:
@@ -1793,6 +1972,11 @@ PreparedQuery PrepareConstraintQuery(ParsedQuery parsed_query, bool in_explicit_
                     storage::kUniqueConstraintsMaxProperties);
                 break;
               case storage::UniqueConstraints::DeletionStatus::NOT_FOUND:
+                constraint_notification.code = NotificationCode::NONEXISTANT_CONSTRAINT;
+                constraint_notification.title =
+                    fmt::format("Constraint UNIQUE on label {} on properties {} doesn't exist.", label_name,
+                                properties_stringified);
+                break;
               case storage::UniqueConstraints::DeletionStatus::SUCCESS:
                 break;
             }
@@ -1804,8 +1988,10 @@ PreparedQuery PrepareConstraintQuery(ParsedQuery parsed_query, bool in_explicit_
 
   return PreparedQuery{{},
                        std::move(parsed_query.required_privileges),
-                       [handler = std::move(handler)](AnyStream *stream, std::optional<int> n) {
-                         handler();
+                       [handler = std::move(handler), constraint_notification = std::move(constraint_notification),
+                        notifications](AnyStream * /*stream*/, std::optional<int> /*n*/) mutable {
+                         handler(constraint_notification);
+                         notifications->push_back(constraint_notification);
                          return QueryHandlerResult::COMMIT;
                        },
                        RWType::NONE};
@@ -1891,6 +2077,7 @@ Interpreter::PrepareResult Interpreter::Prepare(const std::string &query_string,
     if (utils::Downcast<CypherQuery>(parsed_query.query)) {
       prepared_query = PrepareCypherQuery(std::move(parsed_query), &query_execution->summary, interpreter_context_,
                                           &*execution_db_accessor_, &query_execution->execution_memory,
+                                          &query_execution->notifications,
                                           trigger_context_collector_ ? &*trigger_context_collector_ : nullptr);
     } else if (utils::Downcast<ExplainQuery>(parsed_query.query)) {
       prepared_query = PrepareExplainQuery(std::move(parsed_query), &query_execution->summary, interpreter_context_,
@@ -1903,8 +2090,8 @@ Interpreter::PrepareResult Interpreter::Prepare(const std::string &query_string,
       prepared_query = PrepareDumpQuery(std::move(parsed_query), &query_execution->summary, &*execution_db_accessor_,
                                         &query_execution->execution_memory);
     } else if (utils::Downcast<IndexQuery>(parsed_query.query)) {
-      prepared_query = PrepareIndexQuery(std::move(parsed_query), in_explicit_transaction_, &query_execution->summary,
-                                         interpreter_context_, &query_execution->execution_memory_with_exception);
+      prepared_query = PrepareIndexQuery(std::move(parsed_query), in_explicit_transaction_,
+                                         &query_execution->notifications, interpreter_context_);
     } else if (utils::Downcast<AuthQuery>(parsed_query.query)) {
       prepared_query = PrepareAuthQuery(std::move(parsed_query), in_explicit_transaction_, &query_execution->summary,
                                         interpreter_context_, &*execution_db_accessor_,
@@ -1914,23 +2101,25 @@ Interpreter::PrepareResult Interpreter::Prepare(const std::string &query_string,
                                         interpreter_context_, interpreter_context_->db,
                                         &query_execution->execution_memory_with_exception);
     } else if (utils::Downcast<ConstraintQuery>(parsed_query.query)) {
-      prepared_query =
-          PrepareConstraintQuery(std::move(parsed_query), in_explicit_transaction_, &query_execution->summary,
-                                 interpreter_context_, &query_execution->execution_memory_with_exception);
+      prepared_query = PrepareConstraintQuery(std::move(parsed_query), in_explicit_transaction_,
+                                              &query_execution->notifications, interpreter_context_);
     } else if (utils::Downcast<ReplicationQuery>(parsed_query.query)) {
-      prepared_query = PrepareReplicationQuery(std::move(parsed_query), in_explicit_transaction_, interpreter_context_,
-                                               &*execution_db_accessor_);
+      prepared_query =
+          PrepareReplicationQuery(std::move(parsed_query), in_explicit_transaction_, &query_execution->notifications,
+                                  interpreter_context_, &*execution_db_accessor_);
     } else if (utils::Downcast<LockPathQuery>(parsed_query.query)) {
       prepared_query = PrepareLockPathQuery(std::move(parsed_query), in_explicit_transaction_, interpreter_context_,
                                             &*execution_db_accessor_);
     } else if (utils::Downcast<FreeMemoryQuery>(parsed_query.query)) {
       prepared_query = PrepareFreeMemoryQuery(std::move(parsed_query), in_explicit_transaction_, interpreter_context_);
     } else if (utils::Downcast<TriggerQuery>(parsed_query.query)) {
-      prepared_query = PrepareTriggerQuery(std::move(parsed_query), in_explicit_transaction_, interpreter_context_,
-                                           &*execution_db_accessor_, params, username);
+      prepared_query =
+          PrepareTriggerQuery(std::move(parsed_query), in_explicit_transaction_, &query_execution->notifications,
+                              interpreter_context_, &*execution_db_accessor_, params, username);
     } else if (utils::Downcast<StreamQuery>(parsed_query.query)) {
-      prepared_query = PrepareStreamQuery(std::move(parsed_query), in_explicit_transaction_, interpreter_context_,
-                                          &*execution_db_accessor_, params, username);
+      prepared_query =
+          PrepareStreamQuery(std::move(parsed_query), in_explicit_transaction_, &query_execution->notifications,
+                             interpreter_context_, &*execution_db_accessor_, params, username);
     } else if (utils::Downcast<IsolationLevelQuery>(parsed_query.query)) {
       prepared_query =
           PrepareIsolationLevelQuery(std::move(parsed_query), in_explicit_transaction_, interpreter_context_, this);

src/query/interpreter.hpp

@@ -23,10 +23,11 @@
 #include "query/frontend/ast/cypher_main_visitor.hpp"
 #include "query/frontend/stripped.hpp"
 #include "query/interpret/frame.hpp"
+#include "query/metadata.hpp"
 #include "query/plan/operator.hpp"
 #include "query/plan/read_write_type_checker.hpp"
 #include "query/stream.hpp"
-#include "query/streams.hpp"
+#include "query/stream/streams.hpp"
 #include "query/trigger.hpp"
 #include "query/typed_value.hpp"
 #include "storage/v2/isolation_level.hpp"
@@ -164,7 +165,7 @@ struct PreparedQuery {
  */
 struct InterpreterContext {
   explicit InterpreterContext(storage::Storage *db, InterpreterConfig config,
-                              const std::filesystem::path &data_directory, std::string kafka_bootstrap_servers);
+                              const std::filesystem::path &data_directory);
 
   storage::Storage *db;
 
@@ -189,7 +190,7 @@ struct InterpreterContext {
 
   const InterpreterConfig config;
 
-  query::Streams streams;
+  query::stream::Streams streams;
 };
 
 /// Function that is used to tell all active interpreters that they should stop
@@ -285,6 +286,7 @@ class Interpreter final {
     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;
@@ -377,6 +379,14 @@ std::map<std::string, TypedValue> Interpreter::Pull(TStream *result_stream, std:
     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:

src/query/metadata.cpp

@@ -0,0 +1,117 @@
+// 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.
+
+#include "query/metadata.hpp"
+
+#include <algorithm>
+#include <compare>
+#include <string>
+#include <string_view>
+
+namespace query {
+
+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::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_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 query

src/query/metadata.hpp

@@ -0,0 +1,90 @@
+// 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.
+
+#pragma once
+
+#include <cstdint>
+#include <map>
+#include <string>
+#include <string_view>
+#include <type_traits>
+
+#include "query/typed_value.hpp"
+
+namespace query {
+
+enum class SeverityLevel : uint8_t { INFO, WARNING };
+
+enum class NotificationCode : uint8_t {
+  CREATE_CONSTRAINT,
+  CREATE_INDEX,
+  CHECK_STREAM,
+  CREATE_STREAM,
+  CREATE_TRIGGER,
+  DROP_CONSTRAINT,
+  DROP_INDEX,
+  DROP_REPLICA,
+  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 query

src/query/plan/operator.cpp

@@ -12,6 +12,7 @@
 #include "query/plan/operator.hpp"
 
 #include <algorithm>
+#include <cstdint>
 #include <limits>
 #include <queue>
 #include <random>
@@ -171,15 +172,16 @@ CreateNode::CreateNode(const std::shared_ptr<LogicalOperator> &input, const Node
 
 // Creates a vertex on this GraphDb. Returns a reference to vertex placed on the
 // frame.
-VertexAccessor &CreateLocalVertex(const NodeCreationInfo &node_info, Frame *frame, const ExecutionContext &context) {
+VertexAccessor &CreateLocalVertex(const NodeCreationInfo &node_info, Frame *frame, ExecutionContext &context) {
   auto &dba = *context.db_accessor;
   auto new_node = dba.InsertVertex();
+  context.execution_stats[ExecutionStats::Key::CREATED_NODES] += 1;
   for (auto label : node_info.labels) {
     auto maybe_error = new_node.AddLabel(label);
     if (maybe_error.HasError()) {
       switch (maybe_error.GetError()) {
         case storage::Error::SERIALIZATION_ERROR:
-          throw QueryRuntimeException(kSerializationErrorMessage);
+          throw TransactionSerializationException();
         case storage::Error::DELETED_OBJECT:
           throw QueryRuntimeException("Trying to set a label on a deleted node.");
         case storage::Error::VERTEX_HAS_EDGES:
@@ -188,6 +190,7 @@ VertexAccessor &CreateLocalVertex(const NodeCreationInfo &node_info, Frame *fram
           throw QueryRuntimeException("Unexpected error when setting a label.");
       }
     }
+    context.execution_stats[ExecutionStats::Key::CREATED_LABELS] += 1;
   }
   // Evaluator should use the latest accessors, as modified in this query, when
   // setting properties on new nodes.
@@ -295,7 +298,7 @@ EdgeAccessor CreateEdge(const EdgeCreationInfo &edge_info, DbAccessor *dba, Vert
   } else {
     switch (maybe_edge.GetError()) {
       case storage::Error::SERIALIZATION_ERROR:
-        throw QueryRuntimeException(kSerializationErrorMessage);
+        throw TransactionSerializationException();
       case storage::Error::DELETED_OBJECT:
         throw QueryRuntimeException("Trying to create an edge on a deleted node.");
       case storage::Error::VERTEX_HAS_EDGES:
@@ -346,6 +349,7 @@ bool CreateExpand::CreateExpandCursor::Pull(Frame &frame, ExecutionContext &cont
     }
   }();
 
+  context.execution_stats[ExecutionStats::Key::CREATED_EDGES] += 1;
   if (context.trigger_context_collector) {
     context.trigger_context_collector->RegisterCreatedObject(created_edge);
   }
@@ -1917,7 +1921,7 @@ bool Delete::DeleteCursor::Pull(Frame &frame, ExecutionContext &context) {
       if (maybe_value.HasError()) {
         switch (maybe_value.GetError()) {
           case storage::Error::SERIALIZATION_ERROR:
-            throw QueryRuntimeException(kSerializationErrorMessage);
+            throw TransactionSerializationException();
           case storage::Error::DELETED_OBJECT:
           case storage::Error::VERTEX_HAS_EDGES:
           case storage::Error::PROPERTIES_DISABLED:
@@ -1925,7 +1929,7 @@ bool Delete::DeleteCursor::Pull(Frame &frame, ExecutionContext &context) {
             throw QueryRuntimeException("Unexpected error when deleting an edge.");
         }
       }
-
+      context.execution_stats[ExecutionStats::Key::DELETED_EDGES] += 1;
       if (context.trigger_context_collector && maybe_value.GetValue()) {
         context.trigger_context_collector->RegisterDeletedObject(*maybe_value.GetValue());
       }
@@ -1943,7 +1947,7 @@ bool Delete::DeleteCursor::Pull(Frame &frame, ExecutionContext &context) {
           if (res.HasError()) {
             switch (res.GetError()) {
               case storage::Error::SERIALIZATION_ERROR:
-                throw QueryRuntimeException(kSerializationErrorMessage);
+                throw TransactionSerializationException();
               case storage::Error::DELETED_OBJECT:
               case storage::Error::VERTEX_HAS_EDGES:
               case storage::Error::PROPERTIES_DISABLED:
@@ -1952,6 +1956,10 @@ bool Delete::DeleteCursor::Pull(Frame &frame, ExecutionContext &context) {
             }
           }
 
+          context.execution_stats[ExecutionStats::Key::DELETED_NODES] += 1;
+          if (*res) {
+            context.execution_stats[ExecutionStats::Key::DELETED_EDGES] += static_cast<int64_t>((*res)->second.size());
+          }
           std::invoke([&] {
             if (!context.trigger_context_collector || !*res) {
               return;
@@ -1970,7 +1978,7 @@ bool Delete::DeleteCursor::Pull(Frame &frame, ExecutionContext &context) {
           if (res.HasError()) {
             switch (res.GetError()) {
               case storage::Error::SERIALIZATION_ERROR:
-                throw QueryRuntimeException(kSerializationErrorMessage);
+                throw TransactionSerializationException();
               case storage::Error::VERTEX_HAS_EDGES:
                 throw RemoveAttachedVertexException();
               case storage::Error::DELETED_OBJECT:
@@ -1979,7 +1987,7 @@ bool Delete::DeleteCursor::Pull(Frame &frame, ExecutionContext &context) {
                 throw QueryRuntimeException("Unexpected error when deleting a node.");
             }
           }
-
+          context.execution_stats[ExecutionStats::Key::DELETED_NODES] += 1;
           if (context.trigger_context_collector && res.GetValue()) {
             context.trigger_context_collector->RegisterDeletedObject(*res.GetValue());
           }
@@ -2038,7 +2046,7 @@ bool SetProperty::SetPropertyCursor::Pull(Frame &frame, ExecutionContext &contex
   switch (lhs.type()) {
     case TypedValue::Type::Vertex: {
       auto old_value = PropsSetChecked(&lhs.ValueVertex(), self_.property_, rhs);
-
+      context.execution_stats[ExecutionStats::Key::UPDATED_PROPERTIES] += 1;
       if (context.trigger_context_collector) {
         // rhs cannot be moved because it was created with the allocator that is only valid during current pull
         context.trigger_context_collector->RegisterSetObjectProperty(lhs.ValueVertex(), self_.property_,
@@ -2048,7 +2056,7 @@ bool SetProperty::SetPropertyCursor::Pull(Frame &frame, ExecutionContext &contex
     }
     case TypedValue::Type::Edge: {
       auto old_value = PropsSetChecked(&lhs.ValueEdge(), self_.property_, rhs);
-
+      context.execution_stats[ExecutionStats::Key::UPDATED_PROPERTIES] += 1;
       if (context.trigger_context_collector) {
         // rhs cannot be moved because it was created with the allocator that is only valid during current pull
         context.trigger_context_collector->RegisterSetObjectProperty(lhs.ValueEdge(), self_.property_,
@@ -2120,7 +2128,7 @@ void SetPropertiesOnRecord(TRecordAccessor *record, const TypedValue &rhs, SetPr
         case storage::Error::DELETED_OBJECT:
           throw QueryRuntimeException("Trying to set properties on a deleted graph element.");
         case storage::Error::SERIALIZATION_ERROR:
-          throw QueryRuntimeException(kSerializationErrorMessage);
+          throw TransactionSerializationException();
         case storage::Error::PROPERTIES_DISABLED:
           throw QueryRuntimeException("Can't set property because properties on edges are disabled.");
         case storage::Error::VERTEX_HAS_EDGES:
@@ -2176,7 +2184,7 @@ void SetPropertiesOnRecord(TRecordAccessor *record, const TypedValue &rhs, SetPr
           case storage::Error::DELETED_OBJECT:
             throw QueryRuntimeException("Trying to set properties on a deleted graph element.");
           case storage::Error::SERIALIZATION_ERROR:
-            throw QueryRuntimeException(kSerializationErrorMessage);
+            throw TransactionSerializationException();
           case storage::Error::PROPERTIES_DISABLED:
             throw QueryRuntimeException("Can't set property because properties on edges are disabled.");
           case storage::Error::VERTEX_HAS_EDGES:
@@ -2291,7 +2299,7 @@ bool SetLabels::SetLabelsCursor::Pull(Frame &frame, ExecutionContext &context) {
     if (maybe_value.HasError()) {
       switch (maybe_value.GetError()) {
         case storage::Error::SERIALIZATION_ERROR:
-          throw QueryRuntimeException(kSerializationErrorMessage);
+          throw TransactionSerializationException();
         case storage::Error::DELETED_OBJECT:
           throw QueryRuntimeException("Trying to set a label on a deleted node.");
         case storage::Error::VERTEX_HAS_EDGES:
@@ -2349,7 +2357,7 @@ bool RemoveProperty::RemovePropertyCursor::Pull(Frame &frame, ExecutionContext &
         case storage::Error::DELETED_OBJECT:
           throw QueryRuntimeException("Trying to remove a property on a deleted graph element.");
         case storage::Error::SERIALIZATION_ERROR:
-          throw QueryRuntimeException(kSerializationErrorMessage);
+          throw TransactionSerializationException();
         case storage::Error::PROPERTIES_DISABLED:
           throw QueryRuntimeException(
               "Can't remove property because properties on edges are "
@@ -2420,7 +2428,7 @@ bool RemoveLabels::RemoveLabelsCursor::Pull(Frame &frame, ExecutionContext &cont
     if (maybe_value.HasError()) {
       switch (maybe_value.GetError()) {
         case storage::Error::SERIALIZATION_ERROR:
-          throw QueryRuntimeException(kSerializationErrorMessage);
+          throw TransactionSerializationException();
         case storage::Error::DELETED_OBJECT:
           throw QueryRuntimeException("Trying to remove labels from a deleted node.");
         case storage::Error::VERTEX_HAS_EDGES:
@@ -2430,6 +2438,7 @@ bool RemoveLabels::RemoveLabelsCursor::Pull(Frame &frame, ExecutionContext &cont
       }
     }
 
+    context.execution_stats[ExecutionStats::Key::DELETED_LABELS] += 1;
     if (context.trigger_context_collector && *maybe_value) {
       context.trigger_context_collector->RegisterRemovedVertexLabel(vertex, label);
     }

src/query/procedure/mg_procedure_helpers.cpp

@@ -0,0 +1,36 @@
+// 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.
+
+#include "query/procedure/mg_procedure_helpers.hpp"
+
+namespace query::procedure {
+MgpUniquePtr<mgp_value> GetStringValueOrSetError(const char *string, mgp_memory *memory, mgp_result *result) {
+  procedure::MgpUniquePtr<mgp_value> value{nullptr, mgp_value_destroy};
+  const auto success =
+      TryOrSetError([&] { return procedure::CreateMgpObject(value, mgp_value_make_string, string, memory); }, result);
+  if (!success) {
+    value.reset();
+  }
+
+  return value;
+}
+
+bool InsertResultOrSetError(mgp_result *result, mgp_result_record *record, const char *result_name, mgp_value *value) {
+  if (const auto err = mgp_result_record_insert(record, result_name, value); err != MGP_ERROR_NO_ERROR) {
+    const auto error_msg = fmt::format("Unable to set the result for {}, error = {}", result_name, err);
+    static_cast<void>(mgp_result_set_error_msg(result, error_msg.c_str()));
+    return false;
+  }
+
+  return true;
+}
+
+}  // namespace query::procedure

src/query/procedure/mg_procedure_helpers.hpp

@@ -15,6 +15,8 @@
 #include <type_traits>
 #include <utility>
 
+#include <fmt/format.h>
+
 #include "mg_procedure.h"
 
 namespace query::procedure {
@@ -45,4 +47,23 @@ mgp_error CreateMgpObject(MgpUniquePtr<TObj> &obj, TFunc func, TArgs &&...args)
   obj.reset(raw_obj);
   return err;
 }
+
+template <typename Fun>
+[[nodiscard]] bool TryOrSetError(Fun &&func, mgp_result *result) {
+  if (const auto err = func(); err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+    static_cast<void>(mgp_result_set_error_msg(result, "Not enough memory!"));
+    return false;
+  } else if (err != MGP_ERROR_NO_ERROR) {
+    const auto error_msg = fmt::format("Unexpected error ({})!", err);
+    static_cast<void>(mgp_result_set_error_msg(result, error_msg.c_str()));
+    return false;
+  }
+  return true;
+}
+
+[[nodiscard]] MgpUniquePtr<mgp_value> GetStringValueOrSetError(const char *string, mgp_memory *memory,
+                                                               mgp_result *result);
+
+[[nodiscard]] bool InsertResultOrSetError(mgp_result *result, mgp_result_record *record, const char *result_name,
+                                          mgp_value *value);
 }  // namespace query::procedure

src/query/procedure/mg_procedure_impl.cpp

@@ -26,6 +26,7 @@
 #include "module.hpp"
 #include "query/procedure/cypher_types.hpp"
 #include "query/procedure/mg_procedure_helpers.hpp"
+#include "query/stream/common.hpp"
 #include "storage/v2/property_value.hpp"
 #include "storage/v2/view.hpp"
 #include "utils/algorithm.hpp"
@@ -35,6 +36,7 @@
 #include "utils/memory.hpp"
 #include "utils/string.hpp"
 #include "utils/temporal.hpp"
+#include "utils/variant_helpers.hpp"
 
 // This file contains implementation of top level C API functions, but this is
 // all actually part of query::procedure. So use that namespace for simplicity.
@@ -1581,7 +1583,11 @@ mgp_error mgp_vertex_set_property(struct mgp_vertex *v, const char *property_nam
       }
     }
 
-    auto *trigger_ctx_collector = v->graph->ctx->trigger_context_collector;
+    auto &ctx = v->graph->ctx;
+
+    ctx->execution_stats[query::ExecutionStats::Key::UPDATED_PROPERTIES] += 1;
+
+    auto *trigger_ctx_collector = ctx->trigger_context_collector;
     if (!trigger_ctx_collector || !trigger_ctx_collector->ShouldRegisterObjectPropertyChange<query::VertexAccessor>()) {
       return;
     }
@@ -1617,8 +1623,12 @@ mgp_error mgp_vertex_add_label(struct mgp_vertex *v, mgp_label label) {
       }
     }
 
-    if (v->graph->ctx->trigger_context_collector) {
-      v->graph->ctx->trigger_context_collector->RegisterSetVertexLabel(v->impl, label_id);
+    auto &ctx = v->graph->ctx;
+
+    ctx->execution_stats[query::ExecutionStats::Key::CREATED_LABELS] += 1;
+
+    if (ctx->trigger_context_collector) {
+      ctx->trigger_context_collector->RegisterSetVertexLabel(v->impl, label_id);
     }
   });
 }
@@ -1644,8 +1654,13 @@ mgp_error mgp_vertex_remove_label(struct mgp_vertex *v, mgp_label label) {
           throw SerializationException{"Cannot serialize removing a label from a vertex."};
       }
     }
-    if (v->graph->ctx->trigger_context_collector) {
-      v->graph->ctx->trigger_context_collector->RegisterRemovedVertexLabel(v->impl, label_id);
+
+    auto &ctx = v->graph->ctx;
+
+    ctx->execution_stats[query::ExecutionStats::Key::DELETED_LABELS] += 1;
+
+    if (ctx->trigger_context_collector) {
+      ctx->trigger_context_collector->RegisterRemovedVertexLabel(v->impl, label_id);
     }
   });
 }
@@ -1992,6 +2007,10 @@ mgp_error mgp_edge_set_property(struct mgp_edge *e, const char *property_name, m
       }
     }
 
+    auto &ctx = e->from.graph->ctx;
+
+    ctx->execution_stats[query::ExecutionStats::Key::UPDATED_PROPERTIES] += 1;
+
     auto *trigger_ctx_collector = e->from.graph->ctx->trigger_context_collector;
     if (!trigger_ctx_collector || !trigger_ctx_collector->ShouldRegisterObjectPropertyChange<query::EdgeAccessor>()) {
       return;
@@ -2057,8 +2076,12 @@ mgp_error mgp_graph_create_vertex(struct mgp_graph *graph, mgp_memory *memory, m
           throw ImmutableObjectException{"Cannot create a vertex in an immutable graph!"};
         }
         auto vertex = graph->impl->InsertVertex();
-        if (graph->ctx->trigger_context_collector) {
-          graph->ctx->trigger_context_collector->RegisterCreatedObject(vertex);
+
+        auto &ctx = graph->ctx;
+        ctx->execution_stats[query::ExecutionStats::Key::CREATED_NODES] += 1;
+
+        if (ctx->trigger_context_collector) {
+          ctx->trigger_context_collector->RegisterCreatedObject(vertex);
         }
         return NewRawMgpObject<mgp_vertex>(memory, vertex, graph);
       },
@@ -2085,8 +2108,17 @@ mgp_error mgp_graph_delete_vertex(struct mgp_graph *graph, mgp_vertex *vertex) {
           throw SerializationException{"Cannot serialize removing a vertex."};
       }
     }
-    if (graph->ctx->trigger_context_collector && *result) {
-      graph->ctx->trigger_context_collector->RegisterDeletedObject(**result);
+
+    if (!*result) {
+      return;
+    }
+
+    auto &ctx = graph->ctx;
+
+    ctx->execution_stats[query::ExecutionStats::Key::DELETED_NODES] += 1;
+
+    if (ctx->trigger_context_collector) {
+      ctx->trigger_context_collector->RegisterDeletedObject(**result);
     }
   });
 }
@@ -2111,10 +2143,20 @@ mgp_error mgp_graph_detach_delete_vertex(struct mgp_graph *graph, mgp_vertex *ve
       }
     }
 
-    auto *trigger_ctx_collector = graph->ctx->trigger_context_collector;
-    if (!trigger_ctx_collector || !*result) {
+    if (!*result) {
       return;
     }
+
+    auto &ctx = graph->ctx;
+
+    ctx->execution_stats[query::ExecutionStats::Key::DELETED_NODES] += 1;
+    ctx->execution_stats[query::ExecutionStats::Key::DELETED_EDGES] += static_cast<int64_t>((*result)->second.size());
+
+    auto *trigger_ctx_collector = ctx->trigger_context_collector;
+    if (!trigger_ctx_collector) {
+      return;
+    }
+
     trigger_ctx_collector->RegisterDeletedObject((*result)->first);
     if (!trigger_ctx_collector->ShouldRegisterDeletedObject<query::EdgeAccessor>()) {
       return;
@@ -2147,8 +2189,12 @@ mgp_error mgp_graph_create_edge(mgp_graph *graph, mgp_vertex *from, mgp_vertex *
               throw SerializationException{"Cannot serialize creating an edge."};
           }
         }
-        if (graph->ctx->trigger_context_collector) {
-          graph->ctx->trigger_context_collector->RegisterCreatedObject(*edge);
+        auto &ctx = graph->ctx;
+
+        ctx->execution_stats[query::ExecutionStats::Key::CREATED_EDGES] += 1;
+
+        if (ctx->trigger_context_collector) {
+          ctx->trigger_context_collector->RegisterCreatedObject(*edge);
         }
         return NewRawMgpObject<mgp_edge>(memory, edge.GetValue(), from->graph);
       },
@@ -2174,8 +2220,15 @@ mgp_error mgp_graph_delete_edge(struct mgp_graph *graph, mgp_edge *edge) {
           throw SerializationException{"Cannot serialize removing an edge."};
       }
     }
-    if (graph->ctx->trigger_context_collector && *result) {
-      graph->ctx->trigger_context_collector->RegisterDeletedObject(**result);
+
+    if (!*result) {
+      return;
+    }
+    auto &ctx = graph->ctx;
+
+    ctx->execution_stats[query::ExecutionStats::Key::DELETED_EDGES] += 1;
+    if (ctx->trigger_context_collector) {
+      ctx->trigger_context_collector->RegisterDeletedObject(**result);
     }
   });
 }
@@ -2492,28 +2545,133 @@ bool IsValidIdentifierName(const char *name) {
 
 }  // namespace query::procedure
 
+namespace {
+using StreamSourceType = query::stream::StreamSourceType;
+
+class InvalidMessageFunction : public std::invalid_argument {
+ public:
+  InvalidMessageFunction(const StreamSourceType type, const std::string_view function_name)
+      : std::invalid_argument{fmt::format("'{}' is not defined for a message from a stream of type '{}'", function_name,
+                                          StreamSourceTypeToString(type))} {}
+};
+
+StreamSourceType MessageToStreamSourceType(const mgp_message::KafkaMessage & /*msg*/) {
+  return StreamSourceType::KAFKA;
+}
+
+StreamSourceType MessageToStreamSourceType(const mgp_message::PulsarMessage & /*msg*/) {
+  return StreamSourceType::PULSAR;
+}
+
+mgp_source_type StreamSourceTypeToMgpSourceType(const StreamSourceType type) {
+  switch (type) {
+    case StreamSourceType::KAFKA:
+      return mgp_source_type::KAFKA;
+    case StreamSourceType::PULSAR:
+      return mgp_source_type::PULSAR;
+  }
+}
+
+}  // namespace
+
+mgp_error mgp_message_source_type(mgp_message *message, mgp_source_type *result) {
+  return WrapExceptions(
+      [message] {
+        return std::visit(utils::Overloaded{[](const auto &message) {
+                            return StreamSourceTypeToMgpSourceType(MessageToStreamSourceType(message));
+                          }},
+                          message->msg);
+      },
+      result);
+}
+
 mgp_error mgp_message_payload(mgp_message *message, const char **result) {
-  return WrapExceptions([message] { return message->msg->Payload().data(); }, result);
+  return WrapExceptions(
+      [message] {
+        return std::visit(utils::Overloaded{[](const mgp_message::KafkaMessage &msg) { return msg->Payload().data(); },
+                                            [](const mgp_message::PulsarMessage &msg) { return msg.Payload().data(); },
+                                            [](const auto &msg) -> const char * {
+                                              throw InvalidMessageFunction(MessageToStreamSourceType(msg), "payload");
+                                            }},
+                          message->msg);
+      },
+      result);
 }
 
 mgp_error mgp_message_payload_size(mgp_message *message, size_t *result) {
-  return WrapExceptions([message] { return message->msg->Payload().size(); }, result);
+  return WrapExceptions(
+      [message] {
+        return std::visit(utils::Overloaded{[](const mgp_message::KafkaMessage &msg) { return msg->Payload().size(); },
+                                            [](const mgp_message::PulsarMessage &msg) { return msg.Payload().size(); },
+                                            [](const auto &msg) -> size_t {
+                                              throw InvalidMessageFunction(MessageToStreamSourceType(msg),
+                                                                           "payload_size");
+                                            }},
+                          message->msg);
+      },
+      result);
 }
 
 mgp_error mgp_message_topic_name(mgp_message *message, const char **result) {
-  return WrapExceptions([message] { return message->msg->TopicName().data(); }, result);
+  return WrapExceptions(
+      [message] {
+        return std::visit(
+            utils::Overloaded{[](const mgp_message::KafkaMessage &msg) { return msg->TopicName().data(); },
+                              [](const mgp_message::PulsarMessage &msg) { return msg.TopicName().data(); },
+                              [](const auto &msg) -> const char * {
+                                throw InvalidMessageFunction(MessageToStreamSourceType(msg), "topic_name");
+                              }},
+            message->msg);
+      },
+      result);
 }
 
 mgp_error mgp_message_key(mgp_message *message, const char **result) {
-  return WrapExceptions([message] { return message->msg->Key().data(); }, result);
+  return WrapExceptions(
+      [message] {
+        return std::visit(utils::Overloaded{[](const mgp_message::KafkaMessage &msg) { return msg->Key().data(); },
+                                            [](const auto &msg) -> const char * {
+                                              throw InvalidMessageFunction(MessageToStreamSourceType(msg), "key");
+                                            }},
+                          message->msg);
+      },
+      result);
 }
 
 mgp_error mgp_message_key_size(mgp_message *message, size_t *result) {
-  return WrapExceptions([message] { return message->msg->Key().size(); }, result);
+  return WrapExceptions(
+      [message] {
+        return std::visit(utils::Overloaded{[](const mgp_message::KafkaMessage &msg) { return msg->Key().size(); },
+                                            [](const auto &msg) -> size_t {
+                                              throw InvalidMessageFunction(MessageToStreamSourceType(msg), "key_size");
+                                            }},
+                          message->msg);
+      },
+      result);
 }
 
 mgp_error mgp_message_timestamp(mgp_message *message, int64_t *result) {
-  return WrapExceptions([message] { return message->msg->Timestamp(); }, result);
+  return WrapExceptions(
+      [message] {
+        return std::visit(utils::Overloaded{[](const mgp_message::KafkaMessage &msg) { return msg->Timestamp(); },
+                                            [](const auto &msg) -> int64_t {
+                                              throw InvalidMessageFunction(MessageToStreamSourceType(msg), "timestamp");
+                                            }},
+                          message->msg);
+      },
+      result);
+}
+
+mgp_error mgp_message_offset(struct mgp_message *message, int64_t *result) {
+  return WrapExceptions(
+      [message] {
+        return std::visit(utils::Overloaded{[](const mgp_message::KafkaMessage &msg) { return msg->Offset(); },
+                                            [](const auto &msg) -> int64_t {
+                                              throw InvalidMessageFunction(MessageToStreamSourceType(msg), "offset");
+                                            }},
+                          message->msg);
+      },
+      result);
 }
 
 mgp_error mgp_messages_size(mgp_messages *messages, size_t *result) {

src/query/procedure/mg_procedure_impl.hpp

@@ -20,6 +20,7 @@
 #include <ostream>
 
 #include "integrations/kafka/consumer.hpp"
+#include "integrations/pulsar/consumer.hpp"
 #include "query/context.hpp"
 #include "query/db_accessor.hpp"
 #include "query/procedure/cypher_type_ptr.hpp"
@@ -676,6 +677,17 @@ struct mgp_proc {
         results(memory),
         is_write_procedure(is_write_procedure) {}
 
+  /// @throw std::bad_alloc
+  /// @throw std::length_error
+  mgp_proc(const std::string_view name, std::function<void(mgp_list *, mgp_graph *, mgp_result *, mgp_memory *)> cb,
+           utils::MemoryResource *memory, bool is_write_procedure)
+      : name(name, memory),
+        cb(cb),
+        args(memory),
+        opt_args(memory),
+        results(memory),
+        is_write_procedure(is_write_procedure) {}
+
   /// @throw std::bad_alloc
   /// @throw std::length_error
   mgp_proc(const mgp_proc &other, utils::MemoryResource *memory)
@@ -790,7 +802,12 @@ bool IsValidIdentifierName(const char *name);
 }  // namespace query::procedure
 
 struct mgp_message {
-  const integrations::kafka::Message *msg;
+  explicit mgp_message(const integrations::kafka::Message &message) : msg{&message} {}
+  explicit mgp_message(const integrations::pulsar::Message &message) : msg{message} {}
+
+  using KafkaMessage = const integrations::kafka::Message *;
+  using PulsarMessage = integrations::pulsar::Message;
+  std::variant<KafkaMessage, PulsarMessage> msg;
 };
 
 struct mgp_messages {

src/query/procedure/module.cpp

@@ -1,4 +1,4 @@
-// Copyright 2021 Memgraph Ltd.
+// 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
@@ -18,7 +18,9 @@ extern "C" {
 
 #include <optional>
 
-#include "fmt/format.h"
+#include <fmt/format.h>
+#include <unistd.h>
+
 #include "py/py.hpp"
 #include "query/procedure/mg_procedure_helpers.hpp"
 #include "query/procedure/py_module.hpp"
@@ -53,6 +55,8 @@ class BuiltinModule final : public Module {
 
   void AddTransformation(std::string_view name, mgp_trans trans);
 
+  std::optional<std::filesystem::path> Path() const override { return std::nullopt; }
+
  private:
   /// Registered procedures
   std::map<std::string, mgp_proc, std::less<>> procedures_;
@@ -133,6 +137,20 @@ void RegisterMgLoad(ModuleRegistry *module_registry, utils::RWLock *lock, Builti
   module->AddProcedure("load", std::move(load));
 }
 
+namespace {
+[[nodiscard]] bool IsFileEditable(const std::optional<std::filesystem::path> &path) {
+  return path && access(path->c_str(), W_OK) == 0;
+}
+
+std::string GetPathString(const std::optional<std::filesystem::path> &path) {
+  if (!path) {
+    return "builtin";
+  }
+
+  return std::filesystem::canonical(*path).generic_string();
+}
+}  // namespace
+
 void RegisterMgProcedures(
     // We expect modules to be sorted by name.
     const std::map<std::string, std::unique_ptr<Module>, std::less<>> *all_modules, BuiltinModule *module) {
@@ -147,57 +165,80 @@ void RegisterMgProcedures(
       static_assert(
           std::is_same_v<decltype(module->Procedures()), const std::map<std::string, mgp_proc, std::less<>> *>,
           "Expected module procedures to be sorted by name");
+
+      const auto path = module->Path();
+      const auto path_string = GetPathString(path);
+      const auto is_editable = IsFileEditable(path);
+
       for (const auto &[proc_name, proc] : *module->Procedures()) {
         mgp_result_record *record{nullptr};
-        if (const auto err = mgp_result_new_record(result, &record); err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
-          static_cast<void>(mgp_result_set_error_msg(result, "Not enough memory!"));
-          return;
-        } else if (err != MGP_ERROR_NO_ERROR) {
-          static_cast<void>(mgp_result_set_error_msg(result, "Unexpected error"));
+        {
+          const auto success = TryOrSetError([&] { return mgp_result_new_record(result, &record); }, result);
+          if (!success) {
+            return;
+          }
+        }
+
+        const auto path_value = GetStringValueOrSetError(path_string.c_str(), memory, result);
+        if (!path_value) {
           return;
         }
 
+        MgpUniquePtr<mgp_value> is_editable_value{nullptr, mgp_value_destroy};
+        {
+          const auto success = TryOrSetError(
+              [&] { return CreateMgpObject(is_editable_value, mgp_value_make_bool, is_editable, memory); }, result);
+          if (!success) {
+            return;
+          }
+        }
+
         utils::pmr::string full_name(module_name, memory->impl);
         full_name.append(1, '.');
         full_name.append(proc_name);
-        MgpUniquePtr<mgp_value> name_value{nullptr, mgp_value_destroy};
-        if (const auto err = CreateMgpObject(name_value, mgp_value_make_string, full_name.c_str(), memory);
-            err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
-          static_cast<void>(mgp_result_set_error_msg(result, "Not enough memory!"));
-          return;
-        } else if (err != MGP_ERROR_NO_ERROR) {
-          static_cast<void>(mgp_result_set_error_msg(result, "Unexpected error"));
+        const auto name_value = GetStringValueOrSetError(full_name.c_str(), memory, result);
+        if (!name_value) {
           return;
         }
+
         std::stringstream ss;
         ss << module_name << ".";
         PrintProcSignature(proc, &ss);
         const auto signature = ss.str();
-        MgpUniquePtr<mgp_value> signature_value{nullptr, mgp_value_destroy};
-        if (const auto err = CreateMgpObject(signature_value, mgp_value_make_string, signature.c_str(), memory);
-            err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
-          static_cast<void>(mgp_result_set_error_msg(result, "Not enough memory!"));
-          return;
-        } else if (err != MGP_ERROR_NO_ERROR) {
-          static_cast<void>(mgp_result_set_error_msg(result, "Unexpected error"));
+        const auto signature_value = GetStringValueOrSetError(signature.c_str(), memory, result);
+        if (!signature_value) {
           return;
         }
-        MgpUniquePtr<mgp_value> is_write_value{nullptr, mgp_value_destroy};
 
-        if (const auto err =
-                CreateMgpObject(is_write_value, mgp_value_make_bool, proc.is_write_procedure ? 1 : 0, memory);
-            err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
-          static_cast<void>(mgp_result_set_error_msg(result, "Not enough memory!"));
-          return;
-        } else if (err != MGP_ERROR_NO_ERROR) {
-          static_cast<void>(mgp_result_set_error_msg(result, "Unexpected error"));
+        MgpUniquePtr<mgp_value> is_write_value{nullptr, mgp_value_destroy};
+        {
+          const auto success = TryOrSetError(
+              [&, &proc = proc] {
+                return CreateMgpObject(is_write_value, mgp_value_make_bool, proc.is_write_procedure ? 1 : 0, memory);
+              },
+              result);
+          if (!success) {
+            return;
+          }
+        }
+
+        if (!InsertResultOrSetError(result, record, "name", name_value.get())) {
           return;
         }
-        const auto err1 = mgp_result_record_insert(record, "name", name_value.get());
-        const auto err2 = mgp_result_record_insert(record, "signature", signature_value.get());
-        const auto err3 = mgp_result_record_insert(record, "is_write", is_write_value.get());
-        if (err1 != MGP_ERROR_NO_ERROR || err2 != MGP_ERROR_NO_ERROR || err3 != MGP_ERROR_NO_ERROR) {
-          static_cast<void>(mgp_result_set_error_msg(result, "Unable to set the result!"));
+
+        if (!InsertResultOrSetError(result, record, "signature", signature_value.get())) {
+          return;
+        }
+
+        if (!InsertResultOrSetError(result, record, "is_write", is_write_value.get())) {
+          return;
+        }
+
+        if (!InsertResultOrSetError(result, record, "path", path_value.get())) {
+          return;
+        }
+
+        if (!InsertResultOrSetError(result, record, "is_editable", is_editable_value.get())) {
           return;
         }
       }
@@ -207,6 +248,8 @@ void RegisterMgProcedures(
   MG_ASSERT(mgp_proc_add_result(&procedures, "name", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
   MG_ASSERT(mgp_proc_add_result(&procedures, "signature", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
   MG_ASSERT(mgp_proc_add_result(&procedures, "is_write", Call<mgp_type *>(mgp_type_bool)) == MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_result(&procedures, "path", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_result(&procedures, "is_editable", Call<mgp_type *>(mgp_type_bool)) == MGP_ERROR_NO_ERROR);
   module->AddProcedure("procedures", std::move(procedures));
 }
 
@@ -219,32 +262,52 @@ void RegisterMgTransformations(const std::map<std::string, std::unique_ptr<Modul
       static_assert(
           std::is_same_v<decltype(module->Transformations()), const std::map<std::string, mgp_trans, std::less<>> *>,
           "Expected module transformations to be sorted by name");
+
+      const auto path = module->Path();
+      const auto path_string = GetPathString(path);
+      const auto is_editable = IsFileEditable(path);
+
       for (const auto &[trans_name, proc] : *module->Transformations()) {
         mgp_result_record *record{nullptr};
-        if (const auto err = mgp_result_new_record(result, &record); err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
-          static_cast<void>(mgp_result_set_error_msg(result, "Not enough memory!"));
-          return;
-        } else if (err != MGP_ERROR_NO_ERROR) {
-          static_cast<void>(mgp_result_set_error_msg(result, "Unexpected error"));
+        {
+          const auto success = TryOrSetError([&] { return mgp_result_new_record(result, &record); }, result);
+          if (!success) {
+            return;
+          }
+        }
+
+        const auto path_value = GetStringValueOrSetError(path_string.c_str(), memory, result);
+        if (!path_value) {
           return;
         }
 
+        MgpUniquePtr<mgp_value> is_editable_value{nullptr, mgp_value_destroy};
+        {
+          const auto success = TryOrSetError(
+              [&] { return CreateMgpObject(is_editable_value, mgp_value_make_bool, is_editable, memory); }, result);
+          if (!success) {
+            return;
+          }
+        }
+
         utils::pmr::string full_name(module_name, memory->impl);
         full_name.append(1, '.');
         full_name.append(trans_name);
 
-        MgpUniquePtr<mgp_value> name_value{nullptr, mgp_value_destroy};
-        if (const auto err = CreateMgpObject(name_value, mgp_value_make_string, full_name.c_str(), memory);
-            err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
-          static_cast<void>(mgp_result_set_error_msg(result, "Not enough memory!"));
-          return;
-        } else if (err != MGP_ERROR_NO_ERROR) {
-          static_cast<void>(mgp_result_set_error_msg(result, "Unexpected error"));
+        const auto name_value = GetStringValueOrSetError(full_name.c_str(), memory, result);
+        if (!name_value) {
           return;
         }
 
-        if (const auto err = mgp_result_record_insert(record, "name", name_value.get()); err != MGP_ERROR_NO_ERROR) {
-          static_cast<void>(mgp_result_set_error_msg(result, "Unable to set the result!"));
+        if (!InsertResultOrSetError(result, record, "name", name_value.get())) {
+          return;
+        }
+
+        if (!InsertResultOrSetError(result, record, "path", path_value.get())) {
+          return;
+        }
+
+        if (!InsertResultOrSetError(result, record, "is_editable", is_editable_value.get())) {
           return;
         }
       }
@@ -252,6 +315,8 @@ void RegisterMgTransformations(const std::map<std::string, std::unique_ptr<Modul
   };
   mgp_proc procedures("transformations", transformations_cb, utils::NewDeleteResource(), false);
   MG_ASSERT(mgp_proc_add_result(&procedures, "name", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_result(&procedures, "path", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_result(&procedures, "is_editable", Call<mgp_type *>(mgp_type_bool)) == MGP_ERROR_NO_ERROR);
   module->AddProcedure("transformations", std::move(procedures));
 }
 
@@ -297,6 +362,8 @@ class SharedLibraryModule final : public Module {
 
   const std::map<std::string, mgp_trans, std::less<>> *Transformations() const override;
 
+  std::optional<std::filesystem::path> Path() const override { return file_path_; }
+
  private:
   /// Path as requested for loading the module from a library.
   std::filesystem::path file_path_;
@@ -325,15 +392,16 @@ bool SharedLibraryModule::Load(const std::filesystem::path &file_path) {
   dlerror();  // Clear any existing error.
   handle_ = dlopen(file_path.c_str(), RTLD_NOW | RTLD_LOCAL);
   if (!handle_) {
-    // NOLINTNEXTLINE(concurrency-mt-unsafe)
-    spdlog::error(utils::MessageWithLink("Unable to load module {}; {}.", file_path, dlerror(), "https://memgr.ph/modules"));
+    spdlog::error(
+        utils::MessageWithLink("Unable to load module {}; {}.", file_path, dlerror(), "https://memgr.ph/modules"));
     return false;
   }
   // Get required mgp_init_module
   init_fn_ = reinterpret_cast<int (*)(mgp_module *, mgp_memory *)>(dlsym(handle_, "mgp_init_module"));
   char *dl_errored = dlerror();
   if (!init_fn_ || dl_errored) {
-    spdlog::error(utils::MessageWithLink("Unable to load module {}; {}.", file_path, dl_errored, "https://memgr.ph/modules"));
+    spdlog::error(
+        utils::MessageWithLink("Unable to load module {}; {}.", file_path, dl_errored, "https://memgr.ph/modules"));
     dlclose(handle_);
     handle_ = nullptr;
     return false;
@@ -353,8 +421,8 @@ bool SharedLibraryModule::Load(const std::filesystem::path &file_path) {
       return with_error(error);
     }
     for (auto &trans : module_def->transformations) {
-      const bool was_result_added = MgpTransAddFixedResult(&trans.second);
-      if (!was_result_added) {
+      const auto error_code = MgpTransAddFixedResult(&trans.second);
+      if (error_code != MGP_ERROR_NO_ERROR) {
         const auto error =
             fmt::format("Unable to add result to transformation in module {}; add result failed", file_path);
         return with_error(error);
@@ -384,8 +452,8 @@ bool SharedLibraryModule::Close() {
     spdlog::warn("When closing module {}; mgp_shutdown_module returned {}", file_path_, shutdown_res);
   }
   if (dlclose(handle_) != 0) {
-    // NOLINTNEXTLINE(concurrency-mt-unsafe)
-    spdlog::error(utils::MessageWithLink("Failed to close module {}; {}.", file_path_, dlerror(), "https://memgr.ph/modules"));
+    spdlog::error(
+        utils::MessageWithLink("Failed to close module {}; {}.", file_path_, dlerror(), "https://memgr.ph/modules"));
     return false;
   }
   spdlog::info("Closed module {}", file_path_);
@@ -423,6 +491,7 @@ class PythonModule final : public Module {
 
   const std::map<std::string, mgp_proc, std::less<>> *Procedures() const override;
   const std::map<std::string, mgp_trans, std::less<>> *Transformations() const override;
+  std::optional<std::filesystem::path> Path() const override { return file_path_; }
 
  private:
   std::filesystem::path file_path_;
@@ -444,7 +513,8 @@ bool PythonModule::Load(const std::filesystem::path &file_path) {
   auto gil = py::EnsureGIL();
   auto maybe_exc = py::AppendToSysPath(file_path.parent_path().c_str());
   if (maybe_exc) {
-    spdlog::error(utils::MessageWithLink("Unable to load module {}; {}.", file_path, *maybe_exc, "https://memgr.ph/modules"));
+    spdlog::error(
+        utils::MessageWithLink("Unable to load module {}; {}.", file_path, *maybe_exc, "https://memgr.ph/modules"));
     return false;
   }
   bool succ = true;
@@ -469,7 +539,8 @@ bool PythonModule::Load(const std::filesystem::path &file_path) {
     return true;
   }
   auto exc_info = py::FetchError().value();
-  spdlog::error(utils::MessageWithLink("Unable to load module {}; {}.", file_path, exc_info, "https://memgr.ph/modules"));
+  spdlog::error(
+      utils::MessageWithLink("Unable to load module {}; {}.", file_path, exc_info, "https://memgr.ph/modules"));
   return false;
 }
 
@@ -534,7 +605,8 @@ bool ModuleRegistry::RegisterModule(const std::string_view &name, std::unique_pt
   MG_ASSERT(!name.empty(), "Module name cannot be empty");
   MG_ASSERT(module, "Tried to register an invalid module");
   if (modules_.find(name) != modules_.end()) {
-    spdlog::error(utils::MessageWithLink("Unable to overwrite an already loaded module {}.", name, "https://memgr.ph/modules"));
+    spdlog::error(
+        utils::MessageWithLink("Unable to overwrite an already loaded module {}.", name, "https://memgr.ph/modules"));
     return false;
   }
   modules_.emplace(name, std::move(module));
@@ -564,7 +636,8 @@ void ModuleRegistry::SetModulesDirectory(std::vector<std::filesystem::path> modu
 
 bool ModuleRegistry::LoadModuleIfFound(const std::filesystem::path &modules_dir, const std::string_view name) {
   if (!utils::DirExists(modules_dir)) {
-    spdlog::error(utils::MessageWithLink("Module directory {} doesn't exist.", modules_dir, "https://memgr.ph/modules"));
+    spdlog::error(
+        utils::MessageWithLink("Module directory {} doesn't exist.", modules_dir, "https://memgr.ph/modules"));
     return false;
   }
   for (const auto &entry : std::filesystem::directory_iterator(modules_dir)) {
@@ -601,7 +674,8 @@ bool ModuleRegistry::LoadOrReloadModuleFromName(const std::string_view name) {
 void ModuleRegistry::LoadModulesFromDirectory(const std::filesystem::path &modules_dir) {
   if (modules_dir.empty()) return;
   if (!utils::DirExists(modules_dir)) {
-    spdlog::error(utils::MessageWithLink("Module directory {} doesn't exist.", modules_dir, "https://memgr.ph/modules"));
+    spdlog::error(
+        utils::MessageWithLink("Module directory {} doesn't exist.", modules_dir, "https://memgr.ph/modules"));
     return;
   }
   for (const auto &entry : std::filesystem::directory_iterator(modules_dir)) {
@@ -638,6 +712,16 @@ void ModuleRegistry::UnloadAllModules() {
 
 utils::MemoryResource &ModuleRegistry::GetSharedMemoryResource() noexcept { return *shared_; }
 
+bool ModuleRegistry::RegisterMgProcedure(const std::string_view name, mgp_proc proc) {
+  std::unique_lock<utils::RWLock> guard(lock_);
+  if (auto module = modules_.find("mg"); module != modules_.end()) {
+    auto *builtin_module = dynamic_cast<BuiltinModule *>(module->second.get());
+    builtin_module->AddProcedure(name, std::move(proc));
+    return true;
+  }
+  return false;
+}
+
 namespace {
 
 /// This function returns a pair of either
@@ -661,9 +745,9 @@ template <typename T>
 concept ModuleProperties = utils::SameAsAnyOf<T, mgp_proc, mgp_trans>;
 
 template <ModuleProperties T>
-std::optional<std::pair<procedure::ModulePtr, const T *>> MakePairIfPropFound(const ModuleRegistry &module_registry,
-                                                                              std::string_view fully_qualified_name,
-                                                                              utils::MemoryResource *memory) {
+std::optional<std::pair<ModulePtr, const T *>> MakePairIfPropFound(const ModuleRegistry &module_registry,
+                                                                   std::string_view fully_qualified_name,
+                                                                   utils::MemoryResource *memory) {
   auto prop_fun = [](auto &module) {
     if constexpr (std::is_same_v<T, mgp_proc>) {
       return module->Procedures();
@@ -684,13 +768,13 @@ std::optional<std::pair<procedure::ModulePtr, const T *>> MakePairIfPropFound(co
 
 }  // namespace
 
-std::optional<std::pair<procedure::ModulePtr, const mgp_proc *>> FindProcedure(
-    const ModuleRegistry &module_registry, std::string_view fully_qualified_procedure_name,
-    utils::MemoryResource *memory) {
+std::optional<std::pair<ModulePtr, const mgp_proc *>> FindProcedure(const ModuleRegistry &module_registry,
+                                                                    std::string_view fully_qualified_procedure_name,
+                                                                    utils::MemoryResource *memory) {
   return MakePairIfPropFound<mgp_proc>(module_registry, fully_qualified_procedure_name, memory);
 }
 
-std::optional<std::pair<procedure::ModulePtr, const mgp_trans *>> FindTransformation(
+std::optional<std::pair<ModulePtr, const mgp_trans *>> FindTransformation(
     const ModuleRegistry &module_registry, std::string_view fully_qualified_transformation_name,
     utils::MemoryResource *memory) {
   return MakePairIfPropFound<mgp_trans>(module_registry, fully_qualified_transformation_name, memory);

src/query/procedure/module.hpp

@@ -45,6 +45,8 @@ class Module {
   virtual const std::map<std::string, mgp_proc, std::less<>> *Procedures() const = 0;
   /// Returns registered transformations of this module
   virtual const std::map<std::string, mgp_trans, std::less<>> *Transformations() const = 0;
+
+  virtual std::optional<std::filesystem::path> Path() const = 0;
 };
 
 /// Proxy for a registered Module, acquires a read lock from ModuleRegistry.
@@ -117,6 +119,8 @@ class ModuleRegistry final {
   /// Returns the shared memory allocator used by modules
   utils::MemoryResource &GetSharedMemoryResource() noexcept;
 
+  bool RegisterMgProcedure(std::string_view name, mgp_proc proc);
+
  private:
   std::vector<std::filesystem::path> modules_dirs_;
 };

src/query/procedure/py_module.cpp

@@ -19,6 +19,7 @@
 #include <string>
 #include <string_view>
 
+#include "mg_procedure.h"
 #include "query/procedure/mg_procedure_helpers.hpp"
 #include "query/procedure/mg_procedure_impl.hpp"
 #include "utils/memory.hpp"
@@ -180,6 +181,7 @@ PyObject *PyVerticesIteratorGet(PyVerticesIterator *self, PyObject *Py_UNUSED(ig
     return nullptr;
   }
   if (vertex == nullptr) {
+    Py_INCREF(Py_None);
     return Py_None;
   }
   return MakePyVertex(*vertex, self->py_graph);
@@ -194,6 +196,7 @@ PyObject *PyVerticesIteratorNext(PyVerticesIterator *self, PyObject *Py_UNUSED(i
     return nullptr;
   }
   if (vertex == nullptr) {
+    Py_INCREF(Py_None);
     return Py_None;
   }
   return MakePyVertex(*vertex, self->py_graph);
@@ -250,6 +253,7 @@ PyObject *PyEdgesIteratorGet(PyEdgesIterator *self, PyObject *Py_UNUSED(ignored)
     return nullptr;
   }
   if (edge == nullptr) {
+    Py_INCREF(Py_None);
     return Py_None;
   }
   return MakePyEdge(*edge, self->py_graph);
@@ -264,6 +268,7 @@ PyObject *PyEdgesIteratorNext(PyEdgesIterator *self, PyObject *Py_UNUSED(ignored
     return nullptr;
   }
   if (edge == nullptr) {
+    Py_INCREF(Py_None);
     return Py_None;
   }
   return MakePyEdge(*edge, self->py_graph);
@@ -555,6 +560,21 @@ PyObject *PyMessageIsValid(PyMessage *self, PyObject *Py_UNUSED(ignored)) {
   return PyMessagesIsValid(self->messages, nullptr);
 }
 
+PyObject *PyMessageGetSourceType(PyMessage *self, PyObject *Py_UNUSED(ignored)) {
+  MG_ASSERT(self->message);
+  MG_ASSERT(self->memory);
+  mgp_source_type source_type{mgp_source_type::KAFKA};
+  if (RaiseExceptionFromErrorCode(mgp_message_source_type(self->message, &source_type))) {
+    return nullptr;
+  }
+  auto *py_source_type = PyLong_FromLong(static_cast<int64_t>(source_type));
+  if (!py_source_type) {
+    PyErr_SetString(PyExc_RuntimeError, "Unable to get long from source type");
+    return nullptr;
+  }
+  return py_source_type;
+}
+
 PyObject *PyMessageGetPayload(PyMessage *self, PyObject *Py_UNUSED(ignored)) {
   MG_ASSERT(self->message);
   size_t payload_size{0};
@@ -582,7 +602,7 @@ PyObject *PyMessageGetTopicName(PyMessage *self, PyObject *Py_UNUSED(ignored)) {
   }
   auto *py_topic_name = PyUnicode_FromString(topic_name);
   if (!py_topic_name) {
-    PyErr_SetString(PyExc_RuntimeError, "Unable to get raw bytes from payload");
+    PyErr_SetString(PyExc_RuntimeError, "Unable to get string from topic_name");
     return nullptr;
   }
   return py_topic_name;
@@ -622,15 +642,32 @@ PyObject *PyMessageGetTimestamp(PyMessage *self, PyObject *Py_UNUSED(ignored)) {
   return py_int;
 }
 
+PyObject *PyMessageGetOffset(PyMessage *self, PyObject *Py_UNUSED(ignored)) {
+  MG_ASSERT(self->message);
+  MG_ASSERT(self->memory);
+  int64_t offset{0};
+  if (RaiseExceptionFromErrorCode(mgp_message_offset(self->message, &offset))) {
+    return nullptr;
+  }
+  auto *py_int = PyLong_FromLongLong(offset);
+  if (!py_int) {
+    PyErr_SetString(PyExc_IndexError, "Unable to get offset");
+    return nullptr;
+  }
+  return py_int;
+}
+
 // NOLINTNEXTLINE
 static PyMethodDef PyMessageMethods[] = {
     {"__reduce__", reinterpret_cast<PyCFunction>(DisallowPickleAndCopy), METH_NOARGS, "__reduce__ is not supported"},
     {"is_valid", reinterpret_cast<PyCFunction>(PyMessageIsValid), METH_NOARGS,
      "Return True if messages is in valid context and may be used."},
+    {"source_type", reinterpret_cast<PyCFunction>(PyMessageGetSourceType), METH_NOARGS, "Get stream source type."},
     {"payload", reinterpret_cast<PyCFunction>(PyMessageGetPayload), METH_NOARGS, "Get payload"},
     {"topic_name", reinterpret_cast<PyCFunction>(PyMessageGetTopicName), METH_NOARGS, "Get topic name."},
     {"key", reinterpret_cast<PyCFunction>(PyMessageGetKey), METH_NOARGS, "Get message key."},
     {"timestamp", reinterpret_cast<PyCFunction>(PyMessageGetTimestamp), METH_NOARGS, "Get message timestamp."},
+    {"offset", reinterpret_cast<PyCFunction>(PyMessageGetOffset), METH_NOARGS, "Get message offset."},
     {nullptr},
 };
 
@@ -1207,6 +1244,7 @@ PyObject *PyPropertiesIteratorGet(PyPropertiesIterator *self, PyObject *Py_UNUSE
     return nullptr;
   }
   if (property == nullptr) {
+    Py_INCREF(Py_None);
     return Py_None;
   }
   py::Object py_name(PyUnicode_FromString(property->name));
@@ -1225,6 +1263,7 @@ PyObject *PyPropertiesIteratorNext(PyPropertiesIterator *self, PyObject *Py_UNUS
     return nullptr;
   }
   if (property == nullptr) {
+    Py_INCREF(Py_None);
     return Py_None;
   }
   py::Object py_name(PyUnicode_FromString(property->name));
@@ -1905,6 +1944,18 @@ struct PyMgpError {
   const char *docstring;
 };
 
+bool AddModuleConstants(PyObject &module) {
+  // add source type constants
+  if (PyModule_AddIntConstant(&module, "SOURCE_TYPE_KAFKA", static_cast<int64_t>(mgp_source_type::KAFKA))) {
+    return false;
+  }
+  if (PyModule_AddIntConstant(&module, "SOURCE_TYPE_PULSAR", static_cast<int64_t>(mgp_source_type::PULSAR))) {
+    return false;
+  }
+
+  return true;
+}
+
 PyObject *PyInitMgpModule() {
   PyObject *mgp = PyModule_Create(&PyMgpModule);
   if (!mgp) return nullptr;
@@ -1921,6 +1972,9 @@ PyObject *PyInitMgpModule() {
     }
     return true;
   };
+
+  if (!AddModuleConstants(*mgp)) return nullptr;
+
   if (!register_type(&PyPropertiesIteratorType, "PropertiesIterator")) return nullptr;
   if (!register_type(&PyVerticesIteratorType, "VerticesIterator")) return nullptr;
   if (!register_type(&PyEdgesIteratorType, "EdgesIterator")) return nullptr;

src/query/stream/common.cpp

@@ -0,0 +1,45 @@
+// 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.
+
+#include "query/stream/common.hpp"
+
+#include <json/json.hpp>
+
+namespace query::stream {
+namespace {
+const std::string kBatchIntervalKey{"batch_interval"};
+const std::string kBatchSizeKey{"batch_size"};
+const std::string kTransformationName{"transformation_name"};
+}  // namespace
+
+void to_json(nlohmann::json &data, CommonStreamInfo &&common_info) {
+  data[kBatchIntervalKey] = common_info.batch_interval.count();
+  data[kBatchSizeKey] = common_info.batch_size;
+  data[kTransformationName] = common_info.transformation_name;
+}
+
+void from_json(const nlohmann::json &data, CommonStreamInfo &common_info) {
+  if (const auto batch_interval = data.at(kBatchIntervalKey); !batch_interval.is_null()) {
+    using BatchInterval = decltype(common_info.batch_interval);
+    common_info.batch_interval = BatchInterval{batch_interval.get<typename BatchInterval::rep>()};
+  } else {
+    common_info.batch_interval = kDefaultBatchInterval;
+  }
+
+  if (const auto batch_size = data.at(kBatchSizeKey); !batch_size.is_null()) {
+    common_info.batch_size = batch_size.get<decltype(common_info.batch_size)>();
+  } else {
+    common_info.batch_size = kDefaultBatchSize;
+  }
+
+  data.at(kTransformationName).get_to(common_info.transformation_name);
+}
+}  // namespace query::stream

src/query/stream/common.hpp

@@ -0,0 +1,86 @@
+// 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.
+
+#pragma once
+
+#include <chrono>
+#include <cstdint>
+#include <functional>
+#include <optional>
+#include <string>
+
+#include <json/json.hpp>
+
+#include "query/procedure/mg_procedure_impl.hpp"
+
+namespace query::stream {
+
+constexpr std::chrono::milliseconds kDefaultBatchInterval{100};
+constexpr int64_t kDefaultBatchSize{1000};
+
+template <typename TMessage>
+using ConsumerFunction = std::function<void(const std::vector<TMessage> &)>;
+
+struct CommonStreamInfo {
+  std::chrono::milliseconds batch_interval;
+  int64_t batch_size;
+  std::string transformation_name;
+};
+
+template <typename T>
+concept ConvertableToJson = requires(T value, nlohmann::json data) {
+  { to_json(data, std::move(value)) } -> std::same_as<void>;
+  { from_json(data, value) } -> std::same_as<void>;
+};
+
+template <typename T>
+concept ConvertableToMgpMessage = requires(T value) {
+  mgp_message{value};
+};
+
+template <typename TStream>
+concept Stream = requires(TStream stream) {
+  typename TStream::StreamInfo;
+  typename TStream::Message;
+  TStream{std::string{""}, typename TStream::StreamInfo{}, ConsumerFunction<typename TStream::Message>{}};
+  { stream.Start() } -> std::same_as<void>;
+  { stream.Stop() } -> std::same_as<void>;
+  { stream.IsRunning() } -> std::same_as<bool>;
+  {
+    stream.Check(std::optional<std::chrono::milliseconds>{}, std::optional<int64_t>{},
+                 ConsumerFunction<typename TStream::Message>{})
+    } -> std::same_as<void>;
+  requires std::same_as<std::decay_t<decltype(std::declval<typename TStream::StreamInfo>().common_info)>,
+                        CommonStreamInfo>;
+
+  requires ConvertableToMgpMessage<typename TStream::Message>;
+  requires ConvertableToJson<typename TStream::StreamInfo>;
+};
+
+enum class StreamSourceType : uint8_t { KAFKA, PULSAR };
+
+constexpr std::string_view StreamSourceTypeToString(StreamSourceType type) {
+  switch (type) {
+    case StreamSourceType::KAFKA:
+      return "kafka";
+    case StreamSourceType::PULSAR:
+      return "pulsar";
+  }
+}
+
+template <Stream T>
+StreamSourceType StreamType(const T & /*stream*/);
+
+const std::string kCommonInfoKey = "common_info";
+
+void to_json(nlohmann::json &data, CommonStreamInfo &&info);
+void from_json(const nlohmann::json &data, CommonStreamInfo &common_info);
+}  // namespace query::stream

src/query/stream/sources.cpp

@@ -0,0 +1,132 @@
+// 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/stream/sources.hpp"
+
+#include <json/json.hpp>
+
+#include "integrations/constants.hpp"
+
+namespace query::stream {
+KafkaStream::KafkaStream(std::string stream_name, StreamInfo stream_info,
+                         ConsumerFunction<integrations::kafka::Message> consumer_function) {
+  integrations::kafka::ConsumerInfo consumer_info{
+      .consumer_name = std::move(stream_name),
+      .topics = std::move(stream_info.topics),
+      .consumer_group = std::move(stream_info.consumer_group),
+      .bootstrap_servers = std::move(stream_info.bootstrap_servers),
+      .batch_interval = stream_info.common_info.batch_interval,
+      .batch_size = stream_info.common_info.batch_size,
+      .public_configs = std::move(stream_info.configs),
+      .private_configs = std::move(stream_info.credentials),
+  };
+  consumer_.emplace(std::move(consumer_info), std::move(consumer_function));
+};
+
+KafkaStream::StreamInfo KafkaStream::Info(std::string transformation_name) const {
+  const auto &info = consumer_->Info();
+  return {{.batch_interval = info.batch_interval,
+           .batch_size = info.batch_size,
+           .transformation_name = std::move(transformation_name)},
+          .topics = info.topics,
+          .consumer_group = info.consumer_group,
+          .bootstrap_servers = info.bootstrap_servers,
+          .configs = info.public_configs,
+          .credentials = info.private_configs};
+}
+
+void KafkaStream::Start() { consumer_->Start(); }
+void KafkaStream::Stop() { consumer_->Stop(); }
+bool KafkaStream::IsRunning() const { return consumer_->IsRunning(); }
+
+void KafkaStream::Check(std::optional<std::chrono::milliseconds> timeout, std::optional<int64_t> batch_limit,
+                        const ConsumerFunction<integrations::kafka::Message> &consumer_function) const {
+  consumer_->Check(timeout, batch_limit, consumer_function);
+}
+
+utils::BasicResult<std::string> KafkaStream::SetStreamOffset(const int64_t offset) {
+  return consumer_->SetConsumerOffsets(offset);
+}
+
+namespace {
+const std::string kTopicsKey{"topics"};
+const std::string kConsumerGroupKey{"consumer_group"};
+const std::string kBoostrapServers{"bootstrap_servers"};
+const std::string kConfigs{"configs"};
+const std::string kCredentials{"credentials"};
+
+const std::unordered_map<std::string, std::string> kDefaultConfigsMap;
+}  // namespace
+
+void to_json(nlohmann::json &data, KafkaStream::StreamInfo &&info) {
+  data[kCommonInfoKey] = std::move(info.common_info);
+  data[kTopicsKey] = std::move(info.topics);
+  data[kConsumerGroupKey] = info.consumer_group;
+  data[kBoostrapServers] = std::move(info.bootstrap_servers);
+  data[kConfigs] = std::move(info.configs);
+  data[kCredentials] = std::move(info.credentials);
+}
+
+void from_json(const nlohmann::json &data, KafkaStream::StreamInfo &info) {
+  data.at(kCommonInfoKey).get_to(info.common_info);
+  data.at(kTopicsKey).get_to(info.topics);
+  data.at(kConsumerGroupKey).get_to(info.consumer_group);
+  data.at(kBoostrapServers).get_to(info.bootstrap_servers);
+  // These values might not be present in the persisted JSON object
+  info.configs = data.value(kConfigs, kDefaultConfigsMap);
+  info.credentials = data.value(kCredentials, kDefaultConfigsMap);
+}
+
+PulsarStream::PulsarStream(std::string stream_name, StreamInfo stream_info,
+                           ConsumerFunction<integrations::pulsar::Message> consumer_function) {
+  integrations::pulsar::ConsumerInfo consumer_info{.batch_size = stream_info.common_info.batch_size,
+                                                   .batch_interval = stream_info.common_info.batch_interval,
+                                                   .topics = std::move(stream_info.topics),
+                                                   .consumer_name = std::move(stream_name),
+                                                   .service_url = std::move(stream_info.service_url)};
+
+  consumer_.emplace(std::move(consumer_info), std::move(consumer_function));
+};
+
+PulsarStream::StreamInfo PulsarStream::Info(std::string transformation_name) const {
+  const auto &info = consumer_->Info();
+  return {{.batch_interval = info.batch_interval,
+           .batch_size = info.batch_size,
+           .transformation_name = std::move(transformation_name)},
+          .topics = info.topics,
+          .service_url = info.service_url};
+}
+
+void PulsarStream::Start() { consumer_->Start(); }
+void PulsarStream::Stop() { consumer_->Stop(); }
+bool PulsarStream::IsRunning() const { return consumer_->IsRunning(); }
+
+void PulsarStream::Check(std::optional<std::chrono::milliseconds> timeout, std::optional<int64_t> batch_limit,
+                         const ConsumerFunction<Message> &consumer_function) const {
+  consumer_->Check(timeout, batch_limit, consumer_function);
+}
+
+namespace {
+const std::string kServiceUrl{"service_url"};
+}  // namespace
+
+void to_json(nlohmann::json &data, PulsarStream::StreamInfo &&info) {
+  data[kCommonInfoKey] = std::move(info.common_info);
+  data[kTopicsKey] = std::move(info.topics);
+  data[kServiceUrl] = std::move(info.service_url);
+}
+
+void from_json(const nlohmann::json &data, PulsarStream::StreamInfo &info) {
+  data.at(kCommonInfoKey).get_to(info.common_info);
+  data.at(kTopicsKey).get_to(info.topics);
+  data.at(kServiceUrl).get_to(info.service_url);
+}
+}  // namespace query::stream

src/query/stream/sources.hpp

@@ -0,0 +1,93 @@
+// 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/stream/common.hpp"
+
+#include "integrations/kafka/consumer.hpp"
+#include "integrations/pulsar/consumer.hpp"
+
+namespace query::stream {
+
+struct KafkaStream {
+  struct StreamInfo {
+    CommonStreamInfo common_info;
+    std::vector<std::string> topics;
+    std::string consumer_group;
+    std::string bootstrap_servers;
+    std::unordered_map<std::string, std::string> configs;
+    std::unordered_map<std::string, std::string> credentials;
+  };
+
+  using Message = integrations::kafka::Message;
+
+  KafkaStream(std::string stream_name, StreamInfo stream_info,
+              ConsumerFunction<integrations::kafka::Message> consumer_function);
+
+  StreamInfo Info(std::string transformation_name) const;
+
+  void Start();
+  void Stop();
+  bool IsRunning() const;
+
+  void Check(std::optional<std::chrono::milliseconds> timeout, std::optional<int64_t> batch_limit,
+             const ConsumerFunction<Message> &consumer_function) const;
+
+  utils::BasicResult<std::string> SetStreamOffset(int64_t offset);
+
+ private:
+  using Consumer = integrations::kafka::Consumer;
+  std::optional<Consumer> consumer_;
+};
+
+void to_json(nlohmann::json &data, KafkaStream::StreamInfo &&info);
+void from_json(const nlohmann::json &data, KafkaStream::StreamInfo &info);
+
+template <>
+inline StreamSourceType StreamType(const KafkaStream & /*stream*/) {
+  return StreamSourceType::KAFKA;
+}
+
+struct PulsarStream {
+  struct StreamInfo {
+    CommonStreamInfo common_info;
+    std::vector<std::string> topics;
+    std::string service_url;
+  };
+
+  using Message = integrations::pulsar::Message;
+
+  PulsarStream(std::string stream_name, StreamInfo stream_info, ConsumerFunction<Message> consumer_function);
+
+  StreamInfo Info(std::string transformation_name) const;
+
+  void Start();
+  void Stop();
+  bool IsRunning() const;
+
+  void Check(std::optional<std::chrono::milliseconds> timeout, std::optional<int64_t> batch_limit,
+             const ConsumerFunction<Message> &consumer_function) const;
+
+ private:
+  using Consumer = integrations::pulsar::Consumer;
+  std::optional<Consumer> consumer_;
+};
+
+void to_json(nlohmann::json &data, PulsarStream::StreamInfo &&info);
+void from_json(const nlohmann::json &data, PulsarStream::StreamInfo &info);
+
+template <>
+inline StreamSourceType StreamType(const PulsarStream & /*stream*/) {
+  return StreamSourceType::PULSAR;
+}
+
+}  // namespace query::stream

src/query/stream/streams.cpp

@@ -0,0 +1,768 @@
+// 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/stream/streams.hpp"
+
+#include <shared_mutex>
+#include <string_view>
+#include <utility>
+
+#include <spdlog/spdlog.h>
+#include <json/json.hpp>
+
+#include "integrations/constants.hpp"
+#include "mg_procedure.h"
+#include "query/db_accessor.hpp"
+#include "query/discard_value_stream.hpp"
+#include "query/exceptions.hpp"
+#include "query/interpreter.hpp"
+#include "query/procedure/mg_procedure_helpers.hpp"
+#include "query/procedure/mg_procedure_impl.hpp"
+#include "query/procedure/module.hpp"
+#include "query/stream/sources.hpp"
+#include "query/typed_value.hpp"
+#include "utils/event_counter.hpp"
+#include "utils/logging.hpp"
+#include "utils/memory.hpp"
+#include "utils/on_scope_exit.hpp"
+#include "utils/pmr/string.hpp"
+#include "utils/variant_helpers.hpp"
+
+namespace EventCounter {
+extern const Event MessagesConsumed;
+}  // namespace EventCounter
+
+namespace query::stream {
+namespace {
+constexpr auto kExpectedTransformationResultSize = 2;
+const utils::pmr::string query_param_name{"query", utils::NewDeleteResource()};
+const utils::pmr::string params_param_name{"parameters", utils::NewDeleteResource()};
+
+const std::map<std::string, storage::PropertyValue> empty_parameters{};
+
+auto GetStream(auto &map, const std::string &stream_name) {
+  if (auto it = map.find(stream_name); it != map.end()) {
+    return it;
+  }
+  throw StreamsException("Couldn't find stream '{}'", stream_name);
+}
+
+std::pair<TypedValue /*query*/, TypedValue /*parameters*/> ExtractTransformationResult(
+    utils::pmr::map<utils::pmr::string, TypedValue> &&values, const std::string_view transformation_name,
+    const std::string_view stream_name) {
+  if (values.size() != kExpectedTransformationResultSize) {
+    throw StreamsException(
+        "Transformation '{}' in stream '{}' did not yield all fields (query, parameters) as required.",
+        transformation_name, stream_name);
+  }
+
+  auto get_value = [&](const utils::pmr::string &field_name) mutable -> TypedValue & {
+    auto it = values.find(field_name);
+    if (it == values.end()) {
+      throw StreamsException{"Transformation '{}' in stream '{}' did not yield a record with '{}' field.",
+                             transformation_name, stream_name, field_name};
+    };
+    return it->second;
+  };
+
+  auto &query_value = get_value(query_param_name);
+  MG_ASSERT(query_value.IsString());
+  auto &params_value = get_value(params_param_name);
+  MG_ASSERT(params_value.IsNull() || params_value.IsMap());
+  return {std::move(query_value), std::move(params_value)};
+}
+
+template <typename TMessage>
+void CallCustomTransformation(const std::string &transformation_name, const std::vector<TMessage> &messages,
+                              mgp_result &result, storage::Storage::Accessor &storage_accessor,
+                              utils::MemoryResource &memory_resource, const std::string &stream_name) {
+  DbAccessor db_accessor{&storage_accessor};
+  {
+    auto maybe_transformation =
+        procedure::FindTransformation(procedure::gModuleRegistry, transformation_name, utils::NewDeleteResource());
+
+    if (!maybe_transformation) {
+      throw StreamsException("Couldn't find transformation {} for stream '{}'", transformation_name, stream_name);
+    };
+    const auto &trans = *maybe_transformation->second;
+    mgp_messages mgp_messages{mgp_messages::storage_type{&memory_resource}};
+    std::transform(messages.begin(), messages.end(), std::back_inserter(mgp_messages.messages),
+                   [](const TMessage &message) { return mgp_message{message}; });
+    mgp_graph graph{&db_accessor, storage::View::OLD, nullptr};
+    mgp_memory memory{&memory_resource};
+    result.rows.clear();
+    result.error_msg.reset();
+    result.signature = &trans.results;
+
+    MG_ASSERT(result.signature->size() == kExpectedTransformationResultSize);
+    MG_ASSERT(result.signature->contains(query_param_name));
+    MG_ASSERT(result.signature->contains(params_param_name));
+
+    spdlog::trace("Calling transformation in stream '{}'", stream_name);
+    trans.cb(&mgp_messages, &graph, &result, &memory);
+  }
+  if (result.error_msg.has_value()) {
+    throw StreamsException(result.error_msg->c_str());
+  }
+}
+
+template <Stream TStream>
+StreamStatus<TStream> CreateStatus(std::string stream_name, std::string transformation_name,
+                                   std::optional<std::string> owner, const TStream &stream) {
+  return {.name = std::move(stream_name),
+          .type = StreamType(stream),
+          .is_running = stream.IsRunning(),
+          .info = stream.Info(std::move(transformation_name)),
+          .owner = std::move(owner)};
+}
+
+// nlohmann::json doesn't support string_view access yet
+const std::string kStreamName{"name"};
+const std::string kIsRunningKey{"is_running"};
+const std::string kOwner{"owner"};
+const std::string kType{"type"};
+}  // namespace
+
+template <Stream TStream>
+void to_json(nlohmann::json &data, StreamStatus<TStream> &&status) {
+  data[kStreamName] = std::move(status.name);
+  data[kType] = status.type;
+  data[kIsRunningKey] = status.is_running;
+
+  if (status.owner.has_value()) {
+    data[kOwner] = std::move(*status.owner);
+  } else {
+    data[kOwner] = nullptr;
+  }
+
+  to_json(data, std::move(status.info));
+}
+
+template <Stream TStream>
+void from_json(const nlohmann::json &data, StreamStatus<TStream> &status) {
+  data.at(kStreamName).get_to(status.name);
+  data.at(kIsRunningKey).get_to(status.is_running);
+
+  if (const auto &owner = data.at(kOwner); !owner.is_null()) {
+    status.owner = owner.get<typename decltype(status.owner)::value_type>();
+  } else {
+    status.owner = {};
+  }
+
+  from_json(data, status.info);
+}
+
+Streams::Streams(InterpreterContext *interpreter_context, std::filesystem::path directory)
+    : interpreter_context_(interpreter_context), storage_(std::move(directory)) {
+  RegisterProcedures();
+}
+
+void Streams::RegisterProcedures() {
+  RegisterKafkaProcedures();
+  RegisterPulsarProcedures();
+}
+
+void Streams::RegisterKafkaProcedures() {
+  {
+    constexpr std::string_view proc_name = "kafka_set_stream_offset";
+    auto set_stream_offset = [this, proc_name](mgp_list *args, mgp_graph * /*graph*/, mgp_result *result,
+                                               mgp_memory * /*memory*/) {
+      auto *arg_stream_name = procedure::Call<mgp_value *>(mgp_list_at, args, 0);
+      const auto *stream_name = procedure::Call<const char *>(mgp_value_get_string, arg_stream_name);
+      auto *arg_offset = procedure::Call<mgp_value *>(mgp_list_at, args, 1);
+      const auto offset = procedure::Call<int64_t>(mgp_value_get_int, arg_offset);
+      auto lock_ptr = streams_.Lock();
+      auto it = GetStream(*lock_ptr, std::string(stream_name));
+      std::visit(utils::Overloaded{
+                     [&](StreamData<KafkaStream> &kafka_stream) {
+                       auto stream_source_ptr = kafka_stream.stream_source->Lock();
+                       const auto error = stream_source_ptr->SetStreamOffset(offset);
+                       if (error.HasError()) {
+                         MG_ASSERT(mgp_result_set_error_msg(result, error.GetError().c_str()) == MGP_ERROR_NO_ERROR,
+                                   "Unable to set procedure error message of procedure: {}", proc_name);
+                       }
+                     },
+                     [proc_name](auto && /*other*/) {
+                       throw QueryRuntimeException("'{}' can be only used for Kafka stream sources", proc_name);
+                     }},
+                 it->second);
+    };
+
+    mgp_proc proc(proc_name, set_stream_offset, utils::NewDeleteResource(), false);
+    MG_ASSERT(mgp_proc_add_arg(&proc, "stream_name", procedure::Call<mgp_type *>(mgp_type_string)) ==
+              MGP_ERROR_NO_ERROR);
+    MG_ASSERT(mgp_proc_add_arg(&proc, "offset", procedure::Call<mgp_type *>(mgp_type_int)) == MGP_ERROR_NO_ERROR);
+
+    procedure::gModuleRegistry.RegisterMgProcedure(proc_name, std::move(proc));
+  }
+
+  {
+    constexpr std::string_view proc_name = "kafka_stream_info";
+
+    constexpr std::string_view consumer_group_result_name = "consumer_group";
+    constexpr std::string_view topics_result_name = "topics";
+    constexpr std::string_view bootstrap_servers_result_name = "bootstrap_servers";
+    constexpr std::string_view configs_result_name = "configs";
+    constexpr std::string_view credentials_result_name = "credentials";
+
+    auto get_stream_info = [this, proc_name, consumer_group_result_name, topics_result_name,
+                            bootstrap_servers_result_name, configs_result_name, credentials_result_name](
+                               mgp_list *args, mgp_graph * /*graph*/, mgp_result *result, mgp_memory *memory) {
+      auto *arg_stream_name = procedure::Call<mgp_value *>(mgp_list_at, args, 0);
+      const auto *stream_name = procedure::Call<const char *>(mgp_value_get_string, arg_stream_name);
+      auto lock_ptr = streams_.Lock();
+      auto it = GetStream(*lock_ptr, std::string(stream_name));
+      std::visit(
+          utils::Overloaded{
+              [&](StreamData<KafkaStream> &kafka_stream) {
+                auto stream_source_ptr = kafka_stream.stream_source->Lock();
+                const auto info = stream_source_ptr->Info(kafka_stream.transformation_name);
+                mgp_result_record *record{nullptr};
+                {
+                  const auto success =
+                      procedure::TryOrSetError([&] { return mgp_result_new_record(result, &record); }, result);
+                  if (!success) {
+                    return;
+                  }
+                }
+
+                const auto consumer_group_value =
+                    procedure::GetStringValueOrSetError(info.consumer_group.c_str(), memory, result);
+                if (!consumer_group_value) {
+                  return;
+                }
+
+                procedure::MgpUniquePtr<mgp_list> topic_names{nullptr, mgp_list_destroy};
+                {
+                  const auto success = procedure::TryOrSetError(
+                      [&] {
+                        return procedure::CreateMgpObject(topic_names, mgp_list_make_empty, info.topics.size(), memory);
+                      },
+                      result);
+                  if (!success) {
+                    return;
+                  }
+                }
+
+                for (const auto &topic : info.topics) {
+                  auto topic_value = procedure::GetStringValueOrSetError(topic.c_str(), memory, result);
+                  if (!topic_value) {
+                    return;
+                  }
+                  topic_names->elems.push_back(std::move(*topic_value));
+                }
+
+                procedure::MgpUniquePtr<mgp_value> topics_value{nullptr, mgp_value_destroy};
+                {
+                  const auto success = procedure::TryOrSetError(
+                      [&] { return procedure::CreateMgpObject(topics_value, mgp_value_make_list, topic_names.get()); },
+                      result);
+                  if (!success) {
+                    return;
+                  }
+                  static_cast<void>(topic_names.release());
+                }
+
+                const auto bootstrap_servers_value =
+                    procedure::GetStringValueOrSetError(info.bootstrap_servers.c_str(), memory, result);
+                if (!bootstrap_servers_value) {
+                  return;
+                }
+
+                const auto convert_config_map =
+                    [result, memory](const std::unordered_map<std::string, std::string> &configs_to_convert)
+                    -> procedure::MgpUniquePtr<mgp_value> {
+                  procedure::MgpUniquePtr<mgp_value> configs_value{nullptr, mgp_value_destroy};
+                  procedure::MgpUniquePtr<mgp_map> configs{nullptr, mgp_map_destroy};
+                  {
+                    const auto success = procedure::TryOrSetError(
+                        [&] { return procedure::CreateMgpObject(configs, mgp_map_make_empty, memory); }, result);
+                    if (!success) {
+                      return configs_value;
+                    }
+                  }
+
+                  for (const auto &[key, value] : configs_to_convert) {
+                    auto value_value = procedure::GetStringValueOrSetError(value.c_str(), memory, result);
+                    if (!value_value) {
+                      return configs_value;
+                    }
+                    configs->items.emplace(key, std::move(*value_value));
+                  }
+
+                  {
+                    const auto success = procedure::TryOrSetError(
+                        [&] { return procedure::CreateMgpObject(configs_value, mgp_value_make_map, configs.get()); },
+                        result);
+                    if (!success) {
+                      return configs_value;
+                    }
+                    static_cast<void>(configs.release());
+                  }
+                  return configs_value;
+                };
+
+                const auto configs_value = convert_config_map(info.configs);
+                if (configs_value == nullptr) {
+                  return;
+                }
+
+                using CredentialsType = decltype(KafkaStream::StreamInfo::credentials);
+                CredentialsType reducted_credentials;
+                std::transform(info.credentials.begin(), info.credentials.end(),
+                               std::inserter(reducted_credentials, reducted_credentials.end()),
+                               [](const auto &pair) -> CredentialsType::value_type {
+                                 return {pair.first, integrations::kReducted};
+                               });
+
+                const auto credentials_value = convert_config_map(reducted_credentials);
+                if (credentials_value == nullptr) {
+                  return;
+                }
+
+                if (!procedure::InsertResultOrSetError(result, record, consumer_group_result_name.data(),
+                                                       consumer_group_value.get())) {
+                  return;
+                }
+
+                if (!procedure::InsertResultOrSetError(result, record, topics_result_name.data(), topics_value.get())) {
+                  return;
+                }
+
+                if (!procedure::InsertResultOrSetError(result, record, bootstrap_servers_result_name.data(),
+                                                       bootstrap_servers_value.get())) {
+                  return;
+                }
+
+                if (!procedure::InsertResultOrSetError(result, record, configs_result_name.data(),
+                                                       configs_value.get())) {
+                  return;
+                }
+
+                if (!procedure::InsertResultOrSetError(result, record, credentials_result_name.data(),
+                                                       credentials_value.get())) {
+                  return;
+                }
+              },
+              [proc_name](auto && /*other*/) {
+                throw QueryRuntimeException("'{}' can be only used for Kafka stream sources", proc_name);
+              }},
+          it->second);
+    };
+
+    mgp_proc proc(proc_name, get_stream_info, utils::NewDeleteResource(), false);
+    MG_ASSERT(mgp_proc_add_arg(&proc, "stream_name", procedure::Call<mgp_type *>(mgp_type_string)) ==
+              MGP_ERROR_NO_ERROR);
+    MG_ASSERT(mgp_proc_add_result(&proc, consumer_group_result_name.data(),
+                                  procedure::Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
+    MG_ASSERT(
+        mgp_proc_add_result(&proc, topics_result_name.data(),
+                            procedure::Call<mgp_type *>(mgp_type_list, procedure::Call<mgp_type *>(mgp_type_string))) ==
+        MGP_ERROR_NO_ERROR);
+    MG_ASSERT(mgp_proc_add_result(&proc, bootstrap_servers_result_name.data(),
+                                  procedure::Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
+    MG_ASSERT(mgp_proc_add_result(&proc, configs_result_name.data(), procedure::Call<mgp_type *>(mgp_type_map)) ==
+              MGP_ERROR_NO_ERROR);
+    MG_ASSERT(mgp_proc_add_result(&proc, credentials_result_name.data(), procedure::Call<mgp_type *>(mgp_type_map)) ==
+              MGP_ERROR_NO_ERROR);
+
+    procedure::gModuleRegistry.RegisterMgProcedure(proc_name, std::move(proc));
+  }
+}
+
+void Streams::RegisterPulsarProcedures() {
+  {
+    constexpr std::string_view proc_name = "pulsar_stream_info";
+    constexpr std::string_view service_url_result_name = "service_url";
+    constexpr std::string_view topics_result_name = "topics";
+    auto get_stream_info = [this, proc_name, service_url_result_name, topics_result_name](
+                               mgp_list *args, mgp_graph * /*graph*/, mgp_result *result, mgp_memory *memory) {
+      auto *arg_stream_name = procedure::Call<mgp_value *>(mgp_list_at, args, 0);
+      const auto *stream_name = procedure::Call<const char *>(mgp_value_get_string, arg_stream_name);
+      auto lock_ptr = streams_.Lock();
+      auto it = GetStream(*lock_ptr, std::string(stream_name));
+      std::visit(
+          utils::Overloaded{
+              [&](StreamData<PulsarStream> &pulsar_stream) {
+                auto stream_source_ptr = pulsar_stream.stream_source->Lock();
+                const auto info = stream_source_ptr->Info(pulsar_stream.transformation_name);
+                mgp_result_record *record{nullptr};
+                {
+                  const auto success =
+                      procedure::TryOrSetError([&] { return mgp_result_new_record(result, &record); }, result);
+                  if (!success) {
+                    return;
+                  }
+                }
+
+                auto service_url_value = procedure::GetStringValueOrSetError(info.service_url.c_str(), memory, result);
+                if (!service_url_value) {
+                  return;
+                }
+
+                procedure::MgpUniquePtr<mgp_list> topic_names{nullptr, mgp_list_destroy};
+                {
+                  const auto success = procedure::TryOrSetError(
+                      [&] {
+                        return procedure::CreateMgpObject(topic_names, mgp_list_make_empty, info.topics.size(), memory);
+                      },
+                      result);
+                  if (!success) {
+                    return;
+                  }
+                }
+
+                for (const auto &topic : info.topics) {
+                  auto topic_value = procedure::GetStringValueOrSetError(topic.c_str(), memory, result);
+                  if (!topic_value) {
+                    return;
+                  }
+                  topic_names->elems.push_back(std::move(*topic_value));
+                }
+
+                procedure::MgpUniquePtr<mgp_value> topics_value{nullptr, mgp_value_destroy};
+                {
+                  const auto success = procedure::TryOrSetError(
+                      [&] {
+                        return procedure::CreateMgpObject(topics_value, mgp_value_make_list, topic_names.release());
+                      },
+                      result);
+                  if (!success) {
+                    return;
+                  }
+                }
+
+                if (!procedure::InsertResultOrSetError(result, record, topics_result_name.data(), topics_value.get())) {
+                  return;
+                }
+
+                if (!procedure::InsertResultOrSetError(result, record, service_url_result_name.data(),
+                                                       service_url_value.get())) {
+                  return;
+                }
+              },
+              [proc_name](auto && /*other*/) {
+                throw QueryRuntimeException("'{}' can be only used for Pulsar stream sources", proc_name);
+              }},
+          it->second);
+    };
+
+    mgp_proc proc(proc_name, get_stream_info, utils::NewDeleteResource(), false);
+    MG_ASSERT(mgp_proc_add_arg(&proc, "stream_name", procedure::Call<mgp_type *>(mgp_type_string)) ==
+              MGP_ERROR_NO_ERROR);
+    MG_ASSERT(mgp_proc_add_result(&proc, service_url_result_name.data(),
+                                  procedure::Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
+
+    MG_ASSERT(
+        mgp_proc_add_result(&proc, topics_result_name.data(),
+                            procedure::Call<mgp_type *>(mgp_type_list, procedure::Call<mgp_type *>(mgp_type_string))) ==
+        MGP_ERROR_NO_ERROR);
+
+    procedure::gModuleRegistry.RegisterMgProcedure(proc_name, std::move(proc));
+  }
+}
+
+template <Stream TStream>
+void Streams::Create(const std::string &stream_name, typename TStream::StreamInfo info,
+                     std::optional<std::string> owner) {
+  auto locked_streams = streams_.Lock();
+  auto it = CreateConsumer<TStream>(*locked_streams, stream_name, std::move(info), std::move(owner));
+
+  try {
+    std::visit(
+        [&](auto &&stream_data) {
+          const auto stream_source_ptr = stream_data.stream_source->ReadLock();
+          Persist(CreateStatus(stream_name, stream_data.transformation_name, stream_data.owner, *stream_source_ptr));
+        },
+        it->second);
+  } catch (...) {
+    locked_streams->erase(it);
+    throw;
+  }
+}
+
+template void Streams::Create<KafkaStream>(const std::string &stream_name, KafkaStream::StreamInfo info,
+                                           std::optional<std::string> owner);
+template void Streams::Create<PulsarStream>(const std::string &stream_name, PulsarStream::StreamInfo info,
+                                            std::optional<std::string> owner);
+
+template <Stream TStream>
+Streams::StreamsMap::iterator Streams::CreateConsumer(StreamsMap &map, const std::string &stream_name,
+                                                      typename TStream::StreamInfo stream_info,
+                                                      std::optional<std::string> owner) {
+  if (map.contains(stream_name)) {
+    throw StreamsException{"Stream already exists with name '{}'", stream_name};
+  }
+
+  auto *memory_resource = utils::NewDeleteResource();
+
+  auto consumer_function = [interpreter_context = interpreter_context_, memory_resource, stream_name,
+                            transformation_name = stream_info.common_info.transformation_name, owner = owner,
+                            interpreter = std::make_shared<Interpreter>(interpreter_context_),
+                            result = mgp_result{nullptr, memory_resource},
+                            total_retries = interpreter_context_->config.stream_transaction_conflict_retries,
+                            retry_interval = interpreter_context_->config.stream_transaction_retry_interval](
+                               const std::vector<typename TStream::Message> &messages) mutable {
+    auto accessor = interpreter_context->db->Access();
+    EventCounter::IncrementCounter(EventCounter::MessagesConsumed, messages.size());
+    CallCustomTransformation(transformation_name, messages, result, accessor, *memory_resource, stream_name);
+
+    DiscardValueResultStream stream;
+
+    spdlog::trace("Start transaction in stream '{}'", stream_name);
+    utils::OnScopeExit cleanup{[&interpreter, &result]() {
+      result.rows.clear();
+      interpreter->Abort();
+    }};
+
+    const static std::map<std::string, storage::PropertyValue> empty_parameters{};
+    uint32_t i = 0;
+    while (true) {
+      try {
+        interpreter->BeginTransaction();
+        for (auto &row : result.rows) {
+          spdlog::trace("Processing row in stream '{}'", stream_name);
+          auto [query_value, params_value] =
+              ExtractTransformationResult(std::move(row.values), transformation_name, stream_name);
+          storage::PropertyValue params_prop{params_value};
+
+          std::string query{query_value.ValueString()};
+          spdlog::trace("Executing query '{}' in stream '{}'", query, stream_name);
+          auto prepare_result =
+              interpreter->Prepare(query, params_prop.IsNull() ? empty_parameters : params_prop.ValueMap(), nullptr);
+          if (!interpreter_context->auth_checker->IsUserAuthorized(owner, prepare_result.privileges)) {
+            throw StreamsException{
+                "Couldn't execute query '{}' for stream '{}' because the owner is not authorized to execute the "
+                "query!",
+                query, stream_name};
+          }
+          interpreter->PullAll(&stream);
+        }
+
+        spdlog::trace("Commit transaction in stream '{}'", stream_name);
+        interpreter->CommitTransaction();
+        result.rows.clear();
+        break;
+      } catch (const query::TransactionSerializationException &e) {
+        if (i == total_retries) {
+          throw;
+        }
+        ++i;
+        std::this_thread::sleep_for(retry_interval);
+      }
+    }
+  };
+
+  auto insert_result = map.try_emplace(
+      stream_name, StreamData<TStream>{std::move(stream_info.common_info.transformation_name), std::move(owner),
+                                       std::make_unique<SynchronizedStreamSource<TStream>>(
+                                           stream_name, std::move(stream_info), std::move(consumer_function))});
+  MG_ASSERT(insert_result.second, "Unexpected error during storing consumer '{}'", stream_name);
+  return insert_result.first;
+}
+
+void Streams::RestoreStreams() {
+  spdlog::info("Loading streams...");
+  auto locked_streams_map = streams_.Lock();
+  MG_ASSERT(locked_streams_map->empty(), "Cannot restore streams when some streams already exist!");
+
+  for (const auto &[stream_name, stream_data] : storage_) {
+    const auto get_failed_message = [&stream_name = stream_name](const std::string_view message,
+                                                                 const std::string_view nested_message) {
+      return fmt::format("Failed to load stream '{}', because: {} caused by {}", stream_name, message, nested_message);
+    };
+
+    const auto create_consumer = [&, &stream_name = stream_name, this]<typename T>(StreamStatus<T> status,
+                                                                                   auto &&stream_json_data) {
+      try {
+        stream_json_data.get_to(status);
+      } catch (const nlohmann::json::type_error &exception) {
+        spdlog::warn(get_failed_message("invalid type conversion", exception.what()));
+        return;
+      } catch (const nlohmann::json::out_of_range &exception) {
+        spdlog::warn(get_failed_message("non existing field", exception.what()));
+        return;
+      }
+      MG_ASSERT(status.name == stream_name, "Expected stream name is '{}', but got '{}'", status.name, stream_name);
+
+      try {
+        auto it = CreateConsumer<T>(*locked_streams_map, stream_name, std::move(status.info), std::move(status.owner));
+        if (status.is_running) {
+          std::visit(
+              [&](auto &&stream_data) {
+                auto stream_source_ptr = stream_data.stream_source->Lock();
+                stream_source_ptr->Start();
+              },
+              it->second);
+        }
+        spdlog::info("Stream '{}' is loaded", stream_name);
+      } catch (const utils::BasicException &exception) {
+        spdlog::warn(get_failed_message("unexpected error", exception.what()));
+      }
+    };
+
+    auto stream_json_data = nlohmann::json::parse(stream_data);
+    if (const auto it = stream_json_data.find(kType); it != stream_json_data.end()) {
+      const auto stream_type = static_cast<StreamSourceType>(*it);
+      switch (stream_type) {
+        case StreamSourceType::KAFKA:
+          create_consumer(StreamStatus<KafkaStream>{}, std::move(stream_json_data));
+          break;
+        case StreamSourceType::PULSAR:
+          create_consumer(StreamStatus<PulsarStream>{}, std::move(stream_json_data));
+          break;
+      }
+    } else {
+      spdlog::warn(
+          "Unable to load stream '{}', because it does not contain the type of the stream. Most probably the stream "
+          "was saved before Memgraph 2.1. Please recreate the stream manually to make it work. For more information "
+          "please check https://memgraph.com/docs/memgraph/changelog#v210---nov-22-2021 .",
+          stream_json_data.value(kStreamName, "<invalid format>"));
+    }
+  }
+}
+
+void Streams::Drop(const std::string &stream_name) {
+  auto locked_streams = streams_.Lock();
+
+  auto it = GetStream(*locked_streams, stream_name);
+
+  // streams_ is write locked, which means there is no access to it outside of this function, thus only the Test
+  // function can be executing with the consumer, nothing else.
+  // By acquiring the write lock here for the consumer, we make sure there is
+  // no running Test function for this consumer, therefore it can be erased.
+  std::visit([&](auto &&stream_data) { stream_data.stream_source->Lock(); }, it->second);
+
+  locked_streams->erase(it);
+  if (!storage_.Delete(stream_name)) {
+    throw StreamsException("Couldn't delete stream '{}' from persistent store!", stream_name);
+  }
+
+  // TODO(antaljanosbenjamin) Release the transformation
+}
+
+void Streams::Start(const std::string &stream_name) {
+  auto locked_streams = streams_.Lock();
+  auto it = GetStream(*locked_streams, stream_name);
+
+  std::visit(
+      [&, this](auto &&stream_data) {
+        auto stream_source_ptr = stream_data.stream_source->Lock();
+        stream_source_ptr->Start();
+        Persist(CreateStatus(stream_name, stream_data.transformation_name, stream_data.owner, *stream_source_ptr));
+      },
+      it->second);
+}
+
+void Streams::Stop(const std::string &stream_name) {
+  auto locked_streams = streams_.Lock();
+  auto it = GetStream(*locked_streams, stream_name);
+
+  std::visit(
+      [&, this](auto &&stream_data) {
+        auto stream_source_ptr = stream_data.stream_source->Lock();
+        stream_source_ptr->Stop();
+
+        Persist(CreateStatus(stream_name, stream_data.transformation_name, stream_data.owner, *stream_source_ptr));
+      },
+      it->second);
+}
+
+void Streams::StartAll() {
+  for (auto locked_streams = streams_.Lock(); auto &[stream_name, stream_data] : *locked_streams) {
+    std::visit(
+        [&stream_name = stream_name, this](auto &&stream_data) {
+          auto locked_stream_source = stream_data.stream_source->Lock();
+          if (!locked_stream_source->IsRunning()) {
+            locked_stream_source->Start();
+            Persist(
+                CreateStatus(stream_name, stream_data.transformation_name, stream_data.owner, *locked_stream_source));
+          }
+        },
+        stream_data);
+  }
+}
+
+void Streams::StopAll() {
+  for (auto locked_streams = streams_.Lock(); auto &[stream_name, stream_data] : *locked_streams) {
+    std::visit(
+        [&stream_name = stream_name, this](auto &&stream_data) {
+          auto locked_stream_source = stream_data.stream_source->Lock();
+          if (locked_stream_source->IsRunning()) {
+            locked_stream_source->Stop();
+            Persist(
+                CreateStatus(stream_name, stream_data.transformation_name, stream_data.owner, *locked_stream_source));
+          }
+        },
+        stream_data);
+  }
+}
+
+std::vector<StreamStatus<>> Streams::GetStreamInfo() const {
+  std::vector<StreamStatus<>> result;
+  {
+    for (auto locked_streams = streams_.ReadLock(); const auto &[stream_name, stream_data] : *locked_streams) {
+      std::visit(
+          [&, &stream_name = stream_name](auto &&stream_data) {
+            auto locked_stream_source = stream_data.stream_source->ReadLock();
+            auto info = locked_stream_source->Info(stream_data.transformation_name);
+            result.emplace_back(StreamStatus<>{stream_name, StreamType(*locked_stream_source),
+                                               locked_stream_source->IsRunning(), std::move(info.common_info),
+                                               stream_data.owner});
+          },
+          stream_data);
+    }
+  }
+  return result;
+}
+
+TransformationResult Streams::Check(const std::string &stream_name, std::optional<std::chrono::milliseconds> timeout,
+                                    std::optional<int64_t> batch_limit) const {
+  std::optional locked_streams{streams_.ReadLock()};
+  auto it = GetStream(**locked_streams, stream_name);
+
+  return std::visit(
+      [&](auto &&stream_data) {
+        // This depends on the fact that Drop will first acquire a write lock to the consumer, and erase it only after
+        // that
+        const auto locked_stream_source = stream_data.stream_source->ReadLock();
+        const auto transformation_name = stream_data.transformation_name;
+        locked_streams.reset();
+
+        auto *memory_resource = utils::NewDeleteResource();
+        mgp_result result{nullptr, memory_resource};
+        TransformationResult test_result;
+
+        auto consumer_function = [interpreter_context = interpreter_context_, memory_resource, &stream_name,
+                                  &transformation_name = transformation_name, &result,
+                                  &test_result]<typename T>(const std::vector<T> &messages) mutable {
+          auto accessor = interpreter_context->db->Access();
+          CallCustomTransformation(transformation_name, messages, result, accessor, *memory_resource, stream_name);
+
+          for (auto &row : result.rows) {
+            auto [query, parameters] =
+                ExtractTransformationResult(std::move(row.values), transformation_name, stream_name);
+            std::vector<TypedValue> result_row;
+            result_row.reserve(kExpectedTransformationResultSize);
+            result_row.push_back(std::move(query));
+            result_row.push_back(std::move(parameters));
+
+            test_result.push_back(std::move(result_row));
+          }
+        };
+
+        locked_stream_source->Check(timeout, batch_limit, consumer_function);
+        return test_result;
+      },
+      it->second);
+}
+
+}  // namespace query::stream

src/query/stream/streams.hpp

@@ -1,4 +1,4 @@
-// Copyright 2021 Memgraph Ltd.
+// 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
@@ -11,65 +11,77 @@
 
 #pragma once
 
+#include <concepts>
 #include <functional>
 #include <map>
 #include <optional>
+#include <type_traits>
 #include <unordered_map>
 
+#include <json/json.hpp>
+
 #include "integrations/kafka/consumer.hpp"
 #include "kvstore/kvstore.hpp"
+#include "query/stream/common.hpp"
+#include "query/stream/sources.hpp"
 #include "query/typed_value.hpp"
+#include "storage/v2/property_value.hpp"
+#include "utils/event_counter.hpp"
 #include "utils/exceptions.hpp"
 #include "utils/rw_lock.hpp"
 #include "utils/synchronized.hpp"
 
+class StreamsTest;
 namespace query {
 
+struct InterpreterContext;
+
+namespace stream {
+
 class StreamsException : public utils::BasicException {
  public:
   using BasicException::BasicException;
 };
 
-using TransformationResult = std::vector<std::vector<TypedValue>>;
-using TransformFunction = std::function<TransformationResult(const std::vector<integrations::kafka::Message> &)>;
+template <typename T>
+struct StreamInfo;
 
-struct StreamInfo {
-  std::vector<std::string> topics;
-  std::string consumer_group;
-  std::optional<std::chrono::milliseconds> batch_interval;
-  std::optional<int64_t> batch_size;
-  std::string transformation_name;
-  std::optional<std::string> owner;
-  std::string bootstrap_servers;
+template <>
+struct StreamInfo<void> {
+  using Type = CommonStreamInfo;
 };
 
+template <Stream TStream>
+struct StreamInfo<TStream> {
+  using Type = typename TStream::StreamInfo;
+};
+
+template <typename T>
+using StreamInfoType = typename StreamInfo<T>::Type;
+
+template <typename T = void>
 struct StreamStatus {
   std::string name;
-  StreamInfo info;
+  StreamSourceType type;
   bool is_running;
-};
-
-using SynchronizedConsumer = utils::Synchronized<integrations::kafka::Consumer, utils::WritePrioritizedRWLock>;
-
-struct StreamData {
-  std::string transformation_name;
+  StreamInfoType<T> info;
   std::optional<std::string> owner;
-  std::unique_ptr<SynchronizedConsumer> consumer;
 };
 
-struct InterpreterContext;
+using TransformationResult = std::vector<std::vector<TypedValue>>;
 
 /// Manages Kafka consumers.
 ///
 /// This class is responsible for all query supported actions to happen.
 class Streams final {
+  friend StreamsTest;
+
  public:
   /// Initializes the streams.
   ///
   /// @param interpreter_context context to use to run the result of transformations
-  /// @param bootstrap_servers initial list of brokers as a comma separated list of broker host or host:port
   /// @param directory a directory path to store the persisted streams metadata
-  Streams(InterpreterContext *interpreter_context, std::string bootstrap_servers, std::filesystem::path directory);
+  Streams(InterpreterContext *interpreter_context, std::filesystem::path directory);
 
   /// Restores the streams from the persisted metadata.
   /// The restoration is done in a best effort manner, therefore no exception is thrown on failure, but the error is
@@ -85,7 +97,8 @@ class Streams final {
   /// @param stream_info the necessary informations needed to create the Kafka consumer and transform the messages
   ///
   /// @throws StreamsException if the stream with the same name exists or if the creation of Kafka consumer fails
-  void Create(const std::string &stream_name, StreamInfo stream_info);
+  template <Stream TStream>
+  void Create(const std::string &stream_name, typename TStream::StreamInfo info, std::optional<std::string> owner);
 
   /// Deletes an existing stream and all the data that was persisted.
   ///
@@ -123,7 +136,7 @@ class Streams final {
   /// Return current status for all streams.
   /// It might happend that the is_running field is out of date if the one of the streams stops during the invocation of
   /// this function because of an error.
-  std::vector<StreamStatus> GetStreamInfo() const;
+  std::vector<StreamStatus<>> GetStreamInfo() const;
 
   /// Do a dry-run consume from a stream.
   ///
@@ -144,22 +157,41 @@ class Streams final {
   std::string_view BootstrapServers() const;
 
  private:
-  using StreamsMap = std::unordered_map<std::string, StreamData>;
+  template <Stream TStream>
+  using SynchronizedStreamSource = utils::Synchronized<TStream, utils::WritePrioritizedRWLock>;
+
+  template <Stream TStream>
+  struct StreamData {
+    std::string transformation_name;
+    std::optional<std::string> owner;
+    std::unique_ptr<SynchronizedStreamSource<TStream>> stream_source;
+  };
+
+  using StreamDataVariant = std::variant<StreamData<KafkaStream>, StreamData<PulsarStream>>;
+  using StreamsMap = std::unordered_map<std::string, StreamDataVariant>;
   using SynchronizedStreamsMap = utils::Synchronized<StreamsMap, utils::WritePrioritizedRWLock>;
 
-  static StreamStatus CreateStatus(const std::string &name, const std::string &transformation_name,
-                                   const std::optional<std::string> &owner,
-                                   const integrations::kafka::Consumer &consumer);
+  template <Stream TStream>
+  StreamsMap::iterator CreateConsumer(StreamsMap &map, const std::string &stream_name,
+                                      typename TStream::StreamInfo stream_info, std::optional<std::string> owner);
 
-  StreamsMap::iterator CreateConsumer(StreamsMap &map, const std::string &stream_name, StreamInfo stream_info);
+  template <Stream TStream>
+  void Persist(StreamStatus<TStream> &&status) {
+    const std::string stream_name = status.name;
+    if (!storage_.Put(stream_name, nlohmann::json(std::move(status)).dump())) {
+      throw StreamsException{"Couldn't persist steam data for stream '{}'", stream_name};
+    }
+  }
 
-  void Persist(StreamStatus &&status);
+  void RegisterProcedures();
+  void RegisterKafkaProcedures();
+  void RegisterPulsarProcedures();
 
   InterpreterContext *interpreter_context_;
-  std::string bootstrap_servers_;
   kvstore::KVStore storage_;
 
   SynchronizedStreamsMap streams_;
 };
 
+}  // namespace stream
 }  // namespace query

src/query/streams.cpp

@@ -1,437 +0,0 @@
-// 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.
-
-#include "query/streams.hpp"
-
-#include <shared_mutex>
-#include <string_view>
-#include <utility>
-
-#include <spdlog/spdlog.h>
-#include <json/json.hpp>
-#include "query/db_accessor.hpp"
-#include "query/discard_value_stream.hpp"
-#include "query/interpreter.hpp"
-#include "query/procedure/mg_procedure_impl.hpp"
-#include "query/procedure/module.hpp"
-#include "query/typed_value.hpp"
-#include "utils/event_counter.hpp"
-#include "utils/memory.hpp"
-#include "utils/on_scope_exit.hpp"
-#include "utils/pmr/string.hpp"
-
-namespace EventCounter {
-extern const Event MessagesConsumed;
-}  // namespace EventCounter
-
-namespace query {
-
-using Consumer = integrations::kafka::Consumer;
-using ConsumerInfo = integrations::kafka::ConsumerInfo;
-using Message = integrations::kafka::Message;
-namespace {
-constexpr auto kExpectedTransformationResultSize = 2;
-const utils::pmr::string query_param_name{"query", utils::NewDeleteResource()};
-const utils::pmr::string params_param_name{"parameters", utils::NewDeleteResource()};
-const std::map<std::string, storage::PropertyValue> empty_parameters{};
-
-auto GetStream(auto &map, const std::string &stream_name) {
-  if (auto it = map.find(stream_name); it != map.end()) {
-    return it;
-  }
-  throw StreamsException("Couldn't find stream '{}'", stream_name);
-}
-
-void CallCustomTransformation(const std::string &transformation_name, const std::vector<Message> &messages,
-                              mgp_result &result, storage::Storage::Accessor &storage_accessor,
-                              utils::MemoryResource &memory_resource, const std::string &stream_name) {
-  DbAccessor db_accessor{&storage_accessor};
-  {
-    auto maybe_transformation =
-        procedure::FindTransformation(procedure::gModuleRegistry, transformation_name, utils::NewDeleteResource());
-
-    if (!maybe_transformation) {
-      throw StreamsException("Couldn't find transformation {} for stream '{}'", transformation_name, stream_name);
-    };
-    const auto &trans = *maybe_transformation->second;
-    mgp_messages mgp_messages{mgp_messages::storage_type{&memory_resource}};
-    std::transform(messages.begin(), messages.end(), std::back_inserter(mgp_messages.messages),
-                   [](const integrations::kafka::Message &message) { return mgp_message{&message}; });
-    mgp_graph graph{&db_accessor, storage::View::OLD, nullptr};
-    mgp_memory memory{&memory_resource};
-    result.rows.clear();
-    result.error_msg.reset();
-    result.signature = &trans.results;
-
-    MG_ASSERT(result.signature->size() == kExpectedTransformationResultSize);
-    MG_ASSERT(result.signature->contains(query_param_name));
-    MG_ASSERT(result.signature->contains(params_param_name));
-
-    spdlog::trace("Calling transformation in stream '{}'", stream_name);
-    trans.cb(&mgp_messages, &graph, &result, &memory);
-  }
-  if (result.error_msg.has_value()) {
-    throw StreamsException(result.error_msg->c_str());
-  }
-}
-
-std::pair<TypedValue /*query*/, TypedValue /*parameters*/> ExtractTransformationResult(
-    utils::pmr::map<utils::pmr::string, TypedValue> &&values, const std::string_view transformation_name,
-    const std::string_view stream_name) {
-  if (values.size() != kExpectedTransformationResultSize) {
-    throw StreamsException(
-        "Transformation '{}' in stream '{}' did not yield all fields (query, parameters) as required.",
-        transformation_name, stream_name);
-  }
-
-  auto get_value = [&](const utils::pmr::string &field_name) mutable -> TypedValue & {
-    auto it = values.find(field_name);
-    if (it == values.end()) {
-      throw StreamsException{"Transformation '{}' in stream '{}' did not yield a record with '{}' field.",
-                             transformation_name, stream_name, field_name};
-    };
-    return it->second;
-  };
-
-  auto &query_value = get_value(query_param_name);
-  MG_ASSERT(query_value.IsString());
-  auto &params_value = get_value(params_param_name);
-  MG_ASSERT(params_value.IsNull() || params_value.IsMap());
-  return {std::move(query_value), std::move(params_value)};
-}
-}  // namespace
-
-// nlohmann::json doesn't support string_view access yet
-const std::string kStreamName{"name"};
-const std::string kTopicsKey{"topics"};
-const std::string kConsumerGroupKey{"consumer_group"};
-const std::string kBatchIntervalKey{"batch_interval"};
-const std::string kBatchSizeKey{"batch_size"};
-const std::string kIsRunningKey{"is_running"};
-const std::string kTransformationName{"transformation_name"};
-const std::string kOwner{"owner"};
-const std::string kBoostrapServers{"bootstrap_servers"};
-
-void to_json(nlohmann::json &data, StreamStatus &&status) {
-  auto &info = status.info;
-  data[kStreamName] = std::move(status.name);
-  data[kTopicsKey] = std::move(info.topics);
-  data[kConsumerGroupKey] = info.consumer_group;
-
-  if (info.batch_interval) {
-    data[kBatchIntervalKey] = info.batch_interval->count();
-  } else {
-    data[kBatchIntervalKey] = nullptr;
-  }
-
-  if (info.batch_size) {
-    data[kBatchSizeKey] = *info.batch_size;
-  } else {
-    data[kBatchSizeKey] = nullptr;
-  }
-
-  data[kIsRunningKey] = status.is_running;
-  data[kTransformationName] = status.info.transformation_name;
-
-  if (info.owner.has_value()) {
-    data[kOwner] = std::move(*info.owner);
-  } else {
-    data[kOwner] = nullptr;
-  }
-
-  data[kBoostrapServers] = std::move(info.bootstrap_servers);
-}
-
-void from_json(const nlohmann::json &data, StreamStatus &status) {
-  auto &info = status.info;
-  data.at(kStreamName).get_to(status.name);
-  data.at(kTopicsKey).get_to(info.topics);
-  data.at(kConsumerGroupKey).get_to(info.consumer_group);
-
-  if (const auto batch_interval = data.at(kBatchIntervalKey); !batch_interval.is_null()) {
-    using BatchInterval = decltype(info.batch_interval)::value_type;
-    info.batch_interval = BatchInterval{batch_interval.get<BatchInterval::rep>()};
-  } else {
-    info.batch_interval = {};
-  }
-
-  if (const auto batch_size = data.at(kBatchSizeKey); !batch_size.is_null()) {
-    info.batch_size = batch_size.get<decltype(info.batch_size)::value_type>();
-  } else {
-    info.batch_size = {};
-  }
-
-  data.at(kIsRunningKey).get_to(status.is_running);
-  data.at(kTransformationName).get_to(status.info.transformation_name);
-
-  if (const auto &owner = data.at(kOwner); !owner.is_null()) {
-    info.owner = owner.get<decltype(info.owner)::value_type>();
-  } else {
-    info.owner = {};
-  }
-
-  info.owner = data.value(kBoostrapServers, "");
-}
-
-Streams::Streams(InterpreterContext *interpreter_context, std::string bootstrap_servers,
-                 std::filesystem::path directory)
-    : interpreter_context_(interpreter_context),
-      bootstrap_servers_(std::move(bootstrap_servers)),
-      storage_(std::move(directory)) {}
-
-void Streams::RestoreStreams() {
-  spdlog::info("Loading streams...");
-  auto locked_streams_map = streams_.Lock();
-  MG_ASSERT(locked_streams_map->empty(), "Cannot restore streams when some streams already exist!");
-
-  for (const auto &[stream_name, stream_data] : storage_) {
-    const auto get_failed_message = [&stream_name = stream_name](const std::string_view message,
-                                                                 const std::string_view nested_message) {
-      return fmt::format("Failed to load stream '{}', because: {} caused by {}", stream_name, message, nested_message);
-    };
-
-    StreamStatus status;
-    try {
-      nlohmann::json::parse(stream_data).get_to(status);
-    } catch (const nlohmann::json::type_error &exception) {
-      spdlog::warn(get_failed_message("invalid type conversion", exception.what()));
-      continue;
-    } catch (const nlohmann::json::out_of_range &exception) {
-      spdlog::warn(get_failed_message("non existing field", exception.what()));
-      continue;
-    }
-    MG_ASSERT(status.name == stream_name, "Expected stream name is '{}', but got '{}'", status.name, stream_name);
-
-    try {
-      auto it = CreateConsumer(*locked_streams_map, stream_name, std::move(status.info));
-      if (status.is_running) {
-        it->second.consumer->Lock()->Start();
-      }
-      spdlog::info("Stream '{}' is loaded", stream_name);
-    } catch (const utils::BasicException &exception) {
-      spdlog::warn(get_failed_message("unexpected error", exception.what()));
-    }
-  }
-}
-
-void Streams::Create(const std::string &stream_name, StreamInfo info) {
-  auto locked_streams = streams_.Lock();
-  auto it = CreateConsumer(*locked_streams, stream_name, std::move(info));
-
-  try {
-    Persist(
-        CreateStatus(stream_name, it->second.transformation_name, it->second.owner, *it->second.consumer->ReadLock()));
-  } catch (...) {
-    locked_streams->erase(it);
-    throw;
-  }
-}
-
-void Streams::Drop(const std::string &stream_name) {
-  auto locked_streams = streams_.Lock();
-
-  auto it = GetStream(*locked_streams, stream_name);
-
-  // streams_ is write locked, which means there is no access to it outside of this function, thus only the Test
-  // function can be executing with the consumer, nothing else.
-  // By acquiring the write lock here for the consumer, we make sure there is
-  // no running Test function for this consumer, therefore it can be erased.
-  it->second.consumer->Lock();
-  locked_streams->erase(it);
-
-  if (!storage_.Delete(stream_name)) {
-    throw StreamsException("Couldn't delete stream '{}' from persistent store!", stream_name);
-  }
-
-  // TODO(antaljanosbenjamin) Release the transformation
-}
-
-void Streams::Start(const std::string &stream_name) {
-  auto locked_streams = streams_.Lock();
-  auto it = GetStream(*locked_streams, stream_name);
-
-  auto locked_consumer = it->second.consumer->Lock();
-  locked_consumer->Start();
-
-  Persist(CreateStatus(stream_name, it->second.transformation_name, it->second.owner, *locked_consumer));
-}
-
-void Streams::Stop(const std::string &stream_name) {
-  auto locked_streams = streams_.Lock();
-  auto it = GetStream(*locked_streams, stream_name);
-
-  auto locked_consumer = it->second.consumer->Lock();
-  locked_consumer->Stop();
-
-  Persist(CreateStatus(stream_name, it->second.transformation_name, it->second.owner, *locked_consumer));
-}
-
-void Streams::StartAll() {
-  for (auto locked_streams = streams_.Lock(); auto &[stream_name, stream_data] : *locked_streams) {
-    auto locked_consumer = stream_data.consumer->Lock();
-    if (!locked_consumer->IsRunning()) {
-      locked_consumer->Start();
-      Persist(CreateStatus(stream_name, stream_data.transformation_name, stream_data.owner, *locked_consumer));
-    }
-  }
-}
-
-void Streams::StopAll() {
-  for (auto locked_streams = streams_.Lock(); auto &[stream_name, stream_data] : *locked_streams) {
-    auto locked_consumer = stream_data.consumer->Lock();
-    if (locked_consumer->IsRunning()) {
-      locked_consumer->Stop();
-      Persist(CreateStatus(stream_name, stream_data.transformation_name, stream_data.owner, *locked_consumer));
-    }
-  }
-}
-
-std::vector<StreamStatus> Streams::GetStreamInfo() const {
-  std::vector<StreamStatus> result;
-  {
-    for (auto locked_streams = streams_.ReadLock(); const auto &[stream_name, stream_data] : *locked_streams) {
-      result.emplace_back(CreateStatus(stream_name, stream_data.transformation_name, stream_data.owner,
-                                       *stream_data.consumer->ReadLock()));
-    }
-  }
-  return result;
-}
-
-TransformationResult Streams::Check(const std::string &stream_name, std::optional<std::chrono::milliseconds> timeout,
-                                    std::optional<int64_t> batch_limit) const {
-  // This depends on the fact that Drop will first acquire a write lock to the consumer, and erase it only after that
-  auto [locked_consumer,
-        transformation_name] = [this, &stream_name]() -> std::pair<SynchronizedConsumer::ReadLockedPtr, std::string> {
-    auto locked_streams = streams_.ReadLock();
-    auto it = GetStream(*locked_streams, stream_name);
-    return {it->second.consumer->ReadLock(), it->second.transformation_name};
-  }();
-
-  auto *memory_resource = utils::NewDeleteResource();
-  mgp_result result{nullptr, memory_resource};
-  TransformationResult test_result;
-
-  auto consumer_function = [interpreter_context = interpreter_context_, memory_resource, &stream_name,
-                            &transformation_name = transformation_name, &result,
-                            &test_result](const std::vector<Message> &messages) mutable {
-    auto accessor = interpreter_context->db->Access();
-    CallCustomTransformation(transformation_name, messages, result, accessor, *memory_resource, stream_name);
-
-    for (auto &row : result.rows) {
-      auto [query, parameters] = ExtractTransformationResult(std::move(row.values), transformation_name, stream_name);
-      std::vector<TypedValue> result_row;
-      result_row.reserve(kExpectedTransformationResultSize);
-      result_row.push_back(std::move(query));
-      result_row.push_back(std::move(parameters));
-
-      test_result.push_back(std::move(result_row));
-    }
-  };
-
-  locked_consumer->Check(timeout, batch_limit, consumer_function);
-
-  return test_result;
-}
-
-StreamStatus Streams::CreateStatus(const std::string &name, const std::string &transformation_name,
-                                   const std::optional<std::string> &owner,
-                                   const integrations::kafka::Consumer &consumer) {
-  const auto &info = consumer.Info();
-  return StreamStatus{name,
-                      StreamInfo{
-                          info.topics,
-                          info.consumer_group,
-                          info.batch_interval,
-                          info.batch_size,
-                          transformation_name,
-                          owner,
-                      },
-                      consumer.IsRunning()};
-}
-
-Streams::StreamsMap::iterator Streams::CreateConsumer(StreamsMap &map, const std::string &stream_name,
-                                                      StreamInfo stream_info) {
-  if (map.contains(stream_name)) {
-    throw StreamsException{"Stream already exists with name '{}'", stream_name};
-  }
-
-  auto *memory_resource = utils::NewDeleteResource();
-
-  auto consumer_function = [interpreter_context = interpreter_context_, memory_resource, stream_name,
-                            transformation_name = stream_info.transformation_name, owner = stream_info.owner,
-                            interpreter = std::make_shared<Interpreter>(interpreter_context_),
-                            result = mgp_result{nullptr, memory_resource}](
-                               const std::vector<integrations::kafka::Message> &messages) mutable {
-    auto accessor = interpreter_context->db->Access();
-    EventCounter::IncrementCounter(EventCounter::MessagesConsumed, messages.size());
-    CallCustomTransformation(transformation_name, messages, result, accessor, *memory_resource, stream_name);
-
-    DiscardValueResultStream stream;
-
-    spdlog::trace("Start transaction in stream '{}'", stream_name);
-    utils::OnScopeExit cleanup{[&interpreter, &result]() {
-      result.rows.clear();
-      interpreter->Abort();
-    }};
-    interpreter->BeginTransaction();
-
-    for (auto &row : result.rows) {
-      spdlog::trace("Processing row in stream '{}'", stream_name);
-      auto [query_value, params_value] =
-          ExtractTransformationResult(std::move(row.values), transformation_name, stream_name);
-      storage::PropertyValue params_prop{params_value};
-
-      std::string query{query_value.ValueString()};
-      spdlog::trace("Executing query '{}' in stream '{}'", query, stream_name);
-      auto prepare_result =
-          interpreter->Prepare(query, params_prop.IsNull() ? empty_parameters : params_prop.ValueMap(), nullptr);
-      if (!interpreter_context->auth_checker->IsUserAuthorized(owner, prepare_result.privileges)) {
-        throw StreamsException{
-            "Couldn't execute query '{}' for stream '{}' becuase the owner is not authorized to execute the "
-            "query!",
-            query, stream_name};
-      }
-      interpreter->PullAll(&stream);
-    }
-
-    spdlog::trace("Commit transaction in stream '{}'", stream_name);
-    interpreter->CommitTransaction();
-    result.rows.clear();
-  };
-
-  ConsumerInfo consumer_info{
-      .consumer_name = stream_name,
-      .topics = std::move(stream_info.topics),
-      .consumer_group = std::move(stream_info.consumer_group),
-      .batch_interval = stream_info.batch_interval,
-      .batch_size = stream_info.batch_size,
-  };
-
-  auto bootstrap_servers =
-      stream_info.bootstrap_servers.empty() ? bootstrap_servers_ : std::move(stream_info.bootstrap_servers);
-  auto insert_result = map.insert_or_assign(
-      stream_name,
-      StreamData{std::move(stream_info.transformation_name), std::move(stream_info.owner),
-                 std::make_unique<SynchronizedConsumer>(std::move(bootstrap_servers), std::move(consumer_info),
-                                                        std::move(consumer_function))});
-  MG_ASSERT(insert_result.second, "Unexpected error during storing consumer '{}'", stream_name);
-  return insert_result.first;
-}
-
-void Streams::Persist(StreamStatus &&status) {
-  const std::string stream_name = status.name;
-  if (!storage_.Put(stream_name, nlohmann::json(std::move(status)).dump())) {
-    throw StreamsException{"Couldn't persist steam data for stream '{}'", stream_name};
-  }
-}
-
-std::string_view Streams::BootstrapServers() const { return bootstrap_servers_; }
-}  // namespace query

src/requests/CMakeLists.txt

@@ -2,7 +2,10 @@ set(requests_src_files
     requests.cpp)
 
 find_package(CURL REQUIRED)
+find_package(fmt REQUIRED)
+find_package(gflags REQUIRED)
+
 
 add_library(mg-requests STATIC ${requests_src_files})
-target_link_libraries(mg-requests spdlog fmt gflags json ${CURL_LIBRARIES})
+target_link_libraries(mg-requests mg-utils spdlog::spdlog fmt::fmt gflags json ${CURL_LIBRARIES})
 target_include_directories(mg-requests PRIVATE ${CURL_INCLUDE_DIRS})

src/rpc/CMakeLists.txt

@@ -3,6 +3,9 @@ set(rpc_src_files
     protocol.cpp
     server.cpp)
 
+find_package(fmt REQUIRED)
+find_package(gflags REQUIRED)
+
 add_library(mg-rpc STATIC ${rpc_src_files})
-target_link_libraries(mg-rpc Threads::Threads mg-communication mg-utils mg-io fmt gflags)
+target_link_libraries(mg-rpc Threads::Threads mg-communication mg-utils mg-io fmt::fmt gflags)
 target_link_libraries(mg-rpc mg-slk)

src/slk/CMakeLists.txt

@@ -1,6 +1,8 @@
 set(slk_src_files
     streams.cpp)
 
+find_package(gflags REQUIRED)
+
 add_library(mg-slk STATIC ${slk_src_files})
 target_link_libraries(mg-slk gflags)
 target_link_libraries(mg-slk mg-utils)

src/storage/v2/CMakeLists.txt

@@ -30,6 +30,9 @@ set(storage_v2_src_files
 
 #######################
 
+find_package(gflags REQUIRED)
+find_package(Threads REQUIRED)
+
 add_library(mg-storage-v2 STATIC ${storage_v2_src_files})
 target_link_libraries(mg-storage-v2 Threads::Threads mg-utils gflags)
 

src/storage/v2/indices.hpp

@@ -60,6 +60,7 @@ class LabelIndex {
   /// @throw std::bad_alloc
   bool CreateIndex(LabelId label, utils::SkipList<Vertex>::Accessor vertices);
 
+  /// Returns false if there was no index to drop
   bool DropIndex(LabelId label) { return index_.erase(label) > 0; }
 
   bool IndexExists(LabelId label) const { return index_.find(label) != index_.end(); }

src/utils/CMakeLists.txt

@@ -5,11 +5,9 @@ set(utils_src_files
     csv_parsing.cpp
     file.cpp
     file_locker.cpp
-    license.cpp
     memory.cpp
     memory_tracker.cpp
     readable_size.cpp
-    settings.cpp
     signals.cpp
     sysinfo/memory.cpp
     temporal.cpp
@@ -17,5 +15,23 @@ set(utils_src_files
     thread_pool.cpp
     uuid.cpp)
 
+find_package(Boost REQUIRED)
+find_package(fmt REQUIRED)
+find_package(gflags REQUIRED)
+find_package(Threads REQUIRED)
+
 add_library(mg-utils STATIC ${utils_src_files})
-target_link_libraries(mg-utils mg-kvstore mg-slk stdc++fs Threads::Threads spdlog fmt gflags uuid rt)
+target_link_libraries(mg-utils PUBLIC Boost::headers fmt::fmt spdlog::spdlog)
+target_link_libraries(mg-utils PRIVATE stdc++fs Threads::Threads gflags uuid rt)
+
+set(settings_src_files
+    settings.cpp)
+
+add_library(mg-settings STATIC ${settings_src_files})
+target_link_libraries(mg-settings mg-kvstore mg-slk mg-utils)
+
+set(license_src_files
+    license.cpp)
+add_library(mg-license STATIC ${license_src_files})
+target_link_libraries(mg-license mg-settings mg-utils)
+

src/utils/csv_parsing.cpp

@@ -45,7 +45,7 @@ std::optional<utils::pmr::string> Reader::GetNextLine(utils::MemoryResource *mem
 
 Reader::ParsingResult Reader::ParseHeader() {
   // header must be the very first line in the file
-  MG_ASSERT(line_count_ == 1, fmt::format("Invalid use of {}", __func__));
+  MG_ASSERT(line_count_ == 1, "Invalid use of {}", __func__);
   return ParseRow(memory_);
 }
 
@@ -203,7 +203,7 @@ Reader::ParsingResult Reader::ParseRow(utils::MemoryResource *mem) {
   // parse the header.
   // Also, if we don't have a header, the 'number_of_columns_' will be 0, so no
   // need to check the number of columns.
-  if (UNLIKELY(number_of_columns_ != 0 && row.size() != number_of_columns_)) {
+  if (number_of_columns_ != 0 && row.size() != number_of_columns_) [[unlikely]] {
     return ParseError(ParseError::ErrorCode::BAD_NUM_OF_COLUMNS,
                       // ToDo(the-joksim):
                       //    - 'line_count_ - 1' is the last line of a row (as a

src/utils/exceptions.hpp

@@ -38,28 +38,11 @@ namespace utils {
  */
 class BasicException : public std::exception {
  public:
-  /**
-   * @brief Constructor (C strings).
-   *
-   * @param message C-style string error message.
-   *                The string contents are copied upon construction.
-   *                Hence, responsibility for deleting the `char*` lies
-   *                with the caller.
-   */
-  explicit BasicException(const char *message) noexcept : msg_(message) {}
-
   /**
    * @brief Constructor (C++ STL strings).
    *
    * @param message The error message.
    */
-  explicit BasicException(const std::string &message) noexcept : msg_(message) {}
-
-  /**
-   * @brief Constructor (C++ STL string_view).
-   *
-   * @param message The error message.
-   */
   explicit BasicException(const std::string_view message) noexcept : msg_(message) {}
 
   /**
@@ -69,20 +52,8 @@ class BasicException : public std::exception {
    * @param args Arguments for format string.
    */
   template <class... Args>
-  explicit BasicException(const std::string &format, Args &&...args) noexcept
-      : BasicException(fmt::format(format, std::forward<Args>(args)...)) {}
-
-  /**
-   * @brief Constructor with format string (C strings).
-   *
-   * @param format The error format message. The string contents are copied upon
-   * construction. Hence, the responsibility for deleting `char*` lies with the
-   * caller.
-   * @param args Arguments for format string.
-   */
-  template <class... Args>
-  explicit BasicException(const char *format, Args &&...args) noexcept
-      : BasicException(fmt::format(std::string(format), std::forward<Args>(args)...)) {}
+  explicit BasicException(fmt::format_string<Args...> fmt, Args &&...args) noexcept
+      : msg_(fmt::format(fmt, std::forward<Args>(args)...)) {}
 
   /**
    * @brief Virtual destructor to allow for subclassing.
@@ -119,22 +90,12 @@ class BasicException : public std::exception {
  */
 class StacktraceException : public std::exception {
  public:
-  /**
-   * @brief Constructor (C strings).
-   *
-   * @param message C-style string error message.
-   *                The string contents are copied upon construction.
-   *                Hence, responsibility for deleting the `char*` lies
-   *                with the caller.
-   */
-  explicit StacktraceException(const char *message) noexcept : message_(message), stacktrace_(Stacktrace().dump()) {}
-
   /**
    * @brief Constructor (C++ STL strings).
    *
    * @param message The error message.
    */
-  explicit StacktraceException(const std::string &message) noexcept
+  explicit StacktraceException(const std::string_view message) noexcept
       : message_(message), stacktrace_(Stacktrace().dump()) {}
 
   /**
@@ -144,20 +105,8 @@ class StacktraceException : public std::exception {
    * @param args Arguments for format string.
    */
   template <class... Args>
-  explicit StacktraceException(const std::string &format, Args &&...args) noexcept
-      : StacktraceException(fmt::format(format, std::forward<Args>(args)...)) {}
-
-  /**
-   * @brief Constructor with format string (C strings).
-   *
-   * @param format The error format message. The string contents are copied upon
-   * construction. Hence, the responsibility for deleting `char*` lies with the
-   * caller.
-   * @param args Arguments for format string.
-   */
-  template <class... Args>
-  explicit StacktraceException(const char *format, Args &&...args) noexcept
-      : StacktraceException(fmt::format(std::string(format), std::forward<Args>(args)...)) {}
+  explicit StacktraceException(fmt::format_string<Args...> fmt, Args &&...args) noexcept
+      : StacktraceException(fmt::format(fmt, std::forward<Args>(args)...)) {}
 
   /**
    * @brief Virtual destructor to allow for subclassing.
@@ -195,8 +144,8 @@ class NotYetImplemented final : public BasicException {
   explicit NotYetImplemented(const std::string &what) noexcept : BasicException("Not yet implemented: " + what) {}
 
   template <class... Args>
-  explicit NotYetImplemented(const std::string &format, Args &&...args) noexcept
-      : NotYetImplemented(fmt::format(format, std::forward<Args>(args)...)) {}
+  explicit NotYetImplemented(fmt::format_string<Args...> fmt, Args &&...args) noexcept
+      : NotYetImplemented(fmt::format(fmt, std::forward<Args>(args)...)) {}
 };
 
 }  // namespace utils