Dockerfile-analyses

# builds an image for running the notebooks to reproduce analyses and figures
# TODO: might need to do some sort of check to make sure submodules are initialized

FROM debian:bookworm-slim

LABEL maintainer="Paxton Fitzpatrick <Paxton.C.Fitzpatrick.GR@Dartmouth.edu>"

ARG DEBIAN_FRONTEND=noninteractive \
    notebook_ip=0.0.0.0 \
    notebook_port=8888 \
    notebook_dir="/mnt/code/notebooks" \
    develop=0

ENV LANG="C.UTF-8" \
    LC_ALL="C.UTF-8" \
    CONDA_PREFIX="/opt/conda" \
    MAMBA_NO_BANNER=1 \
    PIP_ROOT_USER_ACTION=ignore \
    NOTEBOOK_IP="$notebook_ip" \
    NOTEBOOK_PORT="$notebook_port" \
    NOTEBOOK_DIR="$notebook_dir"

ENV MAMBA_ROOT_PREFIX="$CONDA_PREFIX" \
    PATH="${CONDA_PREFIX}/bin:${PATH}"

SHELL ["/bin/bash", "-o", "pipefail", "-c"]

RUN --mount=type=bind,source=.,target=/mnt,rw \
    sed -i 's/^#force_color_prompt=yes/force_color_prompt=yes/' /etc/skel/.bashrc && \
    # install basic system packages
    apt-get update --fix-missing && \
    apt-get install -y --no-install-recommends eatmydata && \
    eatmydata apt-get install -y --no-install-recommends \
        bzip2 \
        ca-certificates \
        curl \
        git \
        rsync \
        vim-tiny \
        # packages required to build brainiak from source
        # (no pre-built dists available on conda-forge/PyPI for aarch64)
        build-essential \
        libgomp1 \
        libmpich-dev \
        mpich && \
    # symlink "vim.tiny" to just "vim" for simplicity
    ln -s /usr/bin/vim.tiny /usr/bin/vim && \
    # install conda & mamba (faster C++ reimplementation) via miniforge
    curl --progress-bar \
        -L "https://github.com/conda-forge/miniforge/releases/download/25.3.1-0/Miniforge3-25.3.1-0-Linux-$(uname -m).sh" \
        -o "$HOME/miniforge.sh" && \
    /bin/bash "$HOME/miniforge.sh" -bp $CONDA_PREFIX && \
    # configure conda for long-term stability/reproducibility
    conda config --set auto_update_conda false && \
    conda config --set notify_outdated_conda false && \
    conda config --set channel_priority strict && \
    # reconstruct analysis environment from conda-lock file (multi-platform)
    mamba install -yn base -f /mnt/docker/conda-lock.yml && \
    # install & enable GUI menu for configuring Jupyter notebook extensions
    jupyter contrib nbextension install --sys-prefix --overwrite && \
    jupyter nbclassic-extension install --sys-prefix --py jupyter_nbextensions_configurator --overwrite && \
    jupyter nbclassic-extension enable --sys-prefix --py jupyter_nbextensions_configurator && \
    jupyter nbclassic-serverextension enable --sys-prefix --py jupyter_nbextensions_configurator && \
    # copy in Jupyter server & extensions config files from build context
    mkdir -p "$HOME/.jupyter/serverconfig" && \
    cp /mnt/docker/jupyter_notebook_config.py "$HOME/.jupyter/jupyter_notebook_config.py" && \
    cp /mnt/docker/notebook.json "$HOME/.jupyter/serverconfig/notebook.json" && \
    echo -e '{\n  "nbext_hide_incompat": false\n}' > "$HOME/.jupyter/serverconfig/common.json" && \
    # install helpers package for analyses
    if [[ $develop -eq 1 ]]; then editable_flag="-e"; else editable_flag=""; fi && \
    pip install --disable-pip-version-check --no-cache-dir $editable_flag /mnt/code/analysis-helpers && \
    # cleanup to reduce image size
    apt-get clean && \
    apt-get purge -y --auto-remove \
        eatmydata \
        bzip2 \
        ca-certificates \
        curl \
        git \
        build-essential \
        libgomp1 \
        libmpich-dev \
        mpich && \
    rm -rf /var/lib/apt/lists/* && \
    rm "$HOME/miniforge.sh" && \
    mamba clean -afy && \
    find ${CONDA_PREFIX} -follow -type f -name '*.a' -delete && \
    find ${CONDA_PREFIX} -follow -type f -name '*.pyc' -delete && \
    find ${CONDA_PREFIX}/ -follow -type f -name '*.js.map' -delete && \
    rm -rf "$HOME/.cache/pip" && \
    rm -f $HOME/nltk_data/*/*.zip

WORKDIR "/mnt"

ENTRYPOINT ["tini", "-g", "--"]

CMD ["jupyter", "nbclassic"]