basalt 0.2.6

Graph API powered by RocksDB persistent key-value store for fast storage

Basalt - Graph Storage API for C++ and Python

Basalt is a graph storage API powered by RocksDB persistent key-value store for fast storage like NVMe technologies.

Documentation is currently hosted on GitHub: [https://bluebrain.github.io/basalt]

Development stage

The C++ and Python APIs of Basalt are stable already, but substantial additions might come in the future. Thus this library development status is beta.

Usage

C++ API

Conan package

This repository provides a Conan package to ease integration into your existing projects.

CMake

It is also possible to build and install the library using CMake, see build section below.

Python API

Pypi

Python bindings of Basalt are available on Pypi.

Blue Brain 5 supercomputer

Basalt is currently released as module on Blue Brain 5 supercomputer:

$ module purge
$ . /gpfs/bbp.cscs.ch/apps/hpc/jenkins/config/modules.sh
$ module load py-basalt
$ python3
Python 3.6.3 (default, Oct  3 2017, 07:47:49)
[GCC 6.4.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import basalt
>>> basalt.__version__
'0.2.2'
>>> basalt.__rocksdb_version__
'5.17.2'

Manual build and installation instructions

Requirements

• CMake build system, version 3.5.1 or higher.
• RocksDB, a persistent key-value store, version 4.1.1 or higher.
• Python 3, version 3.5 or higher.

Getting the code

This repository grabs a few third-party libraries as git modules. To clone them when you clone basalt, use git clone --recursive option.

If you have already cloned basalt, you can get the git submodules with the command: git submodule update --recursive --init

Building the library 101

C++ Library only

To build the basalt C++ shared library and run the tests:

cd /path/to/basalt
mkdir build
pushd build
cmake ..
CTEST_OUTPUT_ON_FAILURE=1 make all test

To install the library:

pushd build
cmake -DCMAKE_INSTALL_PREFIX=/usr/local .
make all install

Python 3 bindings

To build and run the tests:

cd /path/to/basalt
python3 setup.py test

To install the package:

• with pip: pip3 install -U .
• with distutils: python3 setup.py install
• to create binary tarballs:
  • most simple: python3 setup.py bdist
  • wheel: pip3 install wheel; python3 setup.py bdist_wheel
  • relocatable archive: python3 setup.py bdist_dumb --relative

CMake variables and targets

Main CMake variables:

• Basalt_FORMATTING:BOOL: provide the build target clang-format to check C++ code formatting
• Basalt_STATIC_ANALYSIS:BOOL: provide the build target clang-tidy to perform static analysis of the C++ code
• Basalt_ARCH: value given to the -m compiler option. "native" for instance
• Basalt_PRECOMMIT:BOOL: Enable automatic checks before git commits
• Basalt_CXX_OPTIMIZE:BOOL: Compile C++ with optimization
• Basalt_CXX_SYMBOLS:BOOL: Compile C++ with debug symbols
• Basalt_CXX_WARNINGS:BOOL=ON: Compile C++ with warnings

For a more detailed list, please refer to file CMakeCache.txt in CMake build directory.

CMake targets:

• basalt: build the pure C++ library (without Python bindings)
• _basalt: build the C++ library with Python bindings
• unit-tests: build a C++ executable testing the C++ pure library
• all: build the 3 targets above
• test: execute the tests. It is recommended to execute the command ctest --output-on-failure -VV instead
• install: install the pure C++ library and the CMake configuration required to easily use basalt in another CMake project

Python setuptools commands

Here are the main Python setuptools commands available.

• build: build native library
• test: build and test the package. It also executes the C++ unit-tests as well as the code snippets in the Sphinx documentation.
• install: install the Python package
• doctest: execute the code snippets in the Sphinx documentation
• build_sphinx: build the Sphinx documentation

For instance: python3 setup.py build_sphinx

Files Layout

├── basalt ................... python code of the package
├── cmake
│   └── hpc-coding-conventions git module for C++ code guidelines
├── dev ...................... development related scripts
├── doc ...................... sphinx documentation source code
├── include
│   └── basalt ............... public headers of the C++ library
├── README.md ................ that's me!
├── src
│   ├── basalt ............... C++ library implementation
│   └── third_party .......... C++ libraries (mostly as git modules)
└── tests
    ├── benchmarks ........... scripts to execute before creating a git tag
    ├── py ................... python unit-tests
    └── unit ................. C++ unit-tests using Catch2

Embedded third-parties

External libraries are including either by copy/paste or git submodules in src/third_party directory.

• Catch2: modern, C++-native, header-only, test framework for unit-tests, TDD and BDD unit-test library.
• fmt: A modern formatting library (not part of CMake build yet)
• pybind11: Seamless operability between C++11 and Python
• SpdLog: Fast C++ logging library.

Contributing

If you want to improve the project or you see any issue, every contribution is welcome. Please check contribution guidelines for more information.