Project skeleton for scientific software (C++ with Python interface).
This project skeleton aims to provide a clean and solid basis for small to medium scientific software projects. It is assumed that the software is written in C++ and built as dynamic library in order to provide high performance. An interface for Python is automatically generated. The project skeleton includes several features that facilitate good software engineering practices:
- Usage of a version control system (VCS) and appropriate work flow
- Usage of project management tool including issue tracking
- Automated build system using CMake
- Automated packaging and deployment to a public repository using conda
- Automated testing in continuous integration (CI)
- Automatic documentation generation using Doxygen
- Automatic code formatting checks using clang-format
If you have a conda distribution installed, you can install bertha using
$ conda install -c conda-forge bertha
from the conda-forge channel.
Make sure to clone using
git clone --recurse-submodules [...]
in order to clone the third-party libraries recursively. Alternatively, issue
git submodule update --init --recursive
in the repository if it already exists.
The project is built using the CMake build system, which can be used via GUI or command line. For brevity's sake, only the command line approach is sketched below.
The typical commands under Linux are
$ cd bertha
$ mkdir build
$ cd build
$ cmake ..
$ make -j 4
$ make install
By default, the build type is set to release with debug information, which enables debugging but features the optimization necessary for reasonable performance. Adjust the -j command line parameter of the make command in order to speed up the compilation.
The Doxygen documentation can be build with
$ make doc
and the unit tests can be executed (assuming that CMake has been run with
-DWITH_TESTS) using
$ make test
Using the MinGW environment, the build under Windows is quite similar:
$ cd bertha
$ mkdir build
$ cd build
$ cmake .. -G "Unix Makefiles"
$ make -j 4
$ make install
The only difference is that the generator parameter must be explicitely set to "Unix Makefiles".
Here, you may want to specify a certain version of Visual Studio using
$ cd bertha
$ mkdir build
$ cd build
$ cmake .. -G "Visual Studio 11 2012"
After successful completion there is a Visual Studio .sln file in the build
directory. The build type can be set directly in Visual Studio.
The following software packages are required to build the project:
- CMake >= 3.6
- C++ compiler (any recent version supporting C++11)
- Python >= 2.7
- swig >= 2.0.12
Additionally, the following components are required for unit tests, format checks, documentaton generation, etc.:
- Catch2 (any recent version)
- clang-format (any recent version)
- Doxygen (any recent version)
- gcovr (any recent version)
Feel free to use the code and do not hesitate to create an issue on the project web page. Any contributions are welcome.
If you found the code helpful and used it in your research, you can cite the following paper:
M. Riesch and C. Jirauschek, "bertha: Project skeleton for scientific software.", In preparation, (2019).
If desired, you can cite the project web page directly:
M. Riesch and C. Jirauschek, "bertha: Project skeleton for scientific software.", https://gitlab.com/cph-tum/bertha [Online], (2019).