ci_docker

The CI is configured to:

• run test suite and formatters when there is a merge request on "stable" and "develop" and when there is a commit on "stable".
• build docker and pip package when a tag is pushed.

This was tested in the repo: https://gitlab.synchrotron-soleil.fr/PA/collective-effects/python_cicd

I added a dockerfile based on mine but we can converge toward a standard one. How do you want do deal with optional dependencies like pyPIC & pycollef ? We could do several images per release ?

Dockerfile

-FROM ubuntu:22.04
-RUN useradd -s /bin/bash dockeruser
+FROM gitlab-registry.synchrotron-soleil.fr/pa/collective-effects/python_mpi:latest
 LABEL name mbtrack2
-RUN apt-get update
-RUN apt-get install -y sudo
-RUN adduser dockeruser sudo
-RUN echo '%sudo ALL=(ALL) NOPASSWD:ALL' >> /etc/sudoers
 USER dockeruser
 WORKDIR '/home/dockeruser'
-RUN sudo -E apt-get update
-RUN sudo -E apt-get install -y git && \
-    sudo -E apt-get install -y wget && \
-    sudo -E apt-get install -y gcc g++ &&\
-    sudo -E apt-get install -y gfortran &&\
-    sudo -E apt-get install -y make
-
-RUN wget \
-    https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh \
-    && bash Miniconda3-latest-Linux-x86_64.sh -b \
-    && rm -f Miniconda3-latest-Linux-x86_64.sh \
-    && eval "$(/home/dockeruser/miniconda3/bin/conda shell.bash  hook)"\
-    && conda init
- 
 ENV PATH ${PATH}:/home/dockeruser/miniconda3/bin:/home/dockeruser/miniconda3/condabin
-RUN conda config --add channels conda-forge
-RUN conda install -y gcc_linux-64 gxx_linux-64 gfortran_linux-64
-RUN conda install -y mpi=1.0=openmpi
-RUN conda install -y hdf5=1.14.2=mpi_openmpi_h327c9cf_0
-RUN conda install -y "python=3.11.*" numpy scipy matplotlib mpi4py "h5py>=2.9=mpi*" pandas pip seaborn mpmath
-RUN pip3 install accelerator-toolbox
-
-# Copy and install
+RUN pip3 install accelerator-toolbox==0.5.0
 COPY --chown=dockeruser mbtrack2 /home/dockeruser/mbtrack2
-
 ENV PYTHONPATH=/home/dockeruser/
\ No newline at end of file