Sometimes, we want an interactive C or C++ to test command. Here is the guide to add a C++ kernel to jupyter notebook.

Environments

  • Mac OS X
  • pyenv + virtualenv for a fresh python3 environment.

Cling

Cling is an interactive intepreter for C++. The guide is actually based on cling.

# Download a cling binary
wget https://root.cern.ch/download/cling/cling_2017-03-30_mac1012.tar.bz2

# unzip
tar -xvf cling_2017-03-30_mac1012.tar.bz2

# move to a suitable location
mv cling_2017-03-30_mac1012.tar.bz2 ~/local/cling

# softlink cling
ln -s ~/local/cling/bin/cling ~/bin/cling

You need change the location yourself. After that, cling is included in PATH and we can type

cling --version

to test it.

python

We use pyenv to create a fresh python env.

pyenv global 3.6.0
pyenv virtualenv cpp
pyenv global cpp
pip install jupyter

Now we can type

jupyter notebook

and it should have a python kernel.

C++ Kernel

Cling itself has a jupyter notebook kernel.

cd ~/local/cling/share/cling/Jupyter/kernel
pip install .

# You need choose the version you want.
jupyter kernelspec install cling-cpp14

Now you can use it.

Reference