Dockerfile.base

FROM debian:bookworm

ENV DEBIAN_FRONTEND=noninteractive
ENV GRAMPS_VERSION=60

WORKDIR /app

ENV PYTHONPATH="/usr/lib/python3/dist-packages"

# install poppler (needed for PDF thumbnails)
# ffmpeg (needed for video thumbnails)
# postgresql client (needed for PostgreSQL backend)
RUN apt-get update && apt-get install -y \
  appstream pkg-config libcairo2-dev \
  gir1.2-gtk-3.0 libgirepository1.0-dev libicu-dev \
  graphviz gir1.2-gexiv2-0.10 gir1.2-osmgpsmap-1.0 \
  locales gettext wget python3-pip python3-pil \
  poppler-utils ffmpeg libavcodec-extra \
  unzip \
  libpq-dev postgresql-client postgresql-client-common python3-psycopg2 \
  libgl1-mesa-dev libgtk2.0-dev libatlas-base-dev \
  tesseract-ocr \
  tesseract-ocr-ara \
  tesseract-ocr-bul \
  tesseract-ocr-bre \
  tesseract-ocr-cat \
  tesseract-ocr-ces \
  tesseract-ocr-dan \
  tesseract-ocr-deu \
  tesseract-ocr-ell \
  tesseract-ocr-eng \
  tesseract-ocr-epo \
  tesseract-ocr-spa \
  tesseract-ocr-fin \
  tesseract-ocr-fra \
  tesseract-ocr-gle \
  tesseract-ocr-heb \
  tesseract-ocr-hrv \
  tesseract-ocr-hun \
  tesseract-ocr-isl \
  tesseract-ocr-ind \
  tesseract-ocr-ita \
  tesseract-ocr-jpn \
  tesseract-ocr-kor \
  tesseract-ocr-lit \
  tesseract-ocr-lav \
  tesseract-ocr-mkd \
  tesseract-ocr-nor \
  tesseract-ocr-nld \
  tesseract-ocr-pol \
  tesseract-ocr-por \
  tesseract-ocr-ron \
  tesseract-ocr-rus \
  tesseract-ocr-slk \
  tesseract-ocr-slv \
  tesseract-ocr-sqi \
  tesseract-ocr-srp \
  tesseract-ocr-swe \
  tesseract-ocr-tam \
  tesseract-ocr-tur \
  tesseract-ocr-ukr \
  tesseract-ocr-vie \
  tesseract-ocr-chi-sim \
  tesseract-ocr-chi-tra \
  libopenblas-dev cmake \
  && localedef -i en_US -c -f UTF-8 -A /usr/share/locale/locale.alias en_US.UTF-8 \
  && rm -rf /var/lib/apt/lists/*

# set locale
RUN localedef -i en_US -c -f UTF-8 -A /usr/share/locale/locale.alias en_US.UTF-8
ENV LANGUAGE=en_US.utf8
ENV LANG=en_US.utf8
ENV LC_ALL=en_US.utf8

# limit pytorch to 1 thread
ENV OMP_NUM_THREADS=1

# install wheel first to enable binary packages for all pip installs
RUN python3 -m pip install --break-system-packages --no-cache-dir wheel

# install gunicorn
RUN python3 -m pip install --break-system-packages --no-cache-dir \
    gunicorn

# Install PyICU (slow to compile on ARM, so pre-install in base image)
RUN python3 -m pip install --break-system-packages --no-cache-dir \
    PyICU

# Install PyTorch and opencv
RUN python3 -m pip install --break-system-packages --no-cache-dir --index-url https://download.pytorch.org/whl/cpu torch
RUN python3 -m pip install --break-system-packages --no-cache-dir \
    opencv-python opencv-contrib-python

# Install AI dependencies
RUN python3 -m pip install --break-system-packages --no-cache-dir \
    'sentence-transformers>=4.1.0' \
    'accelerate' \
    'pydantic-ai[openai]>=1.0.0,<2.0.0'

# download and cache sentence transformer model
RUN python3 -c "from sentence_transformers import SentenceTransformer; \
model = SentenceTransformer('sentence-transformers/distiluse-base-multilingual-cased-v2');"

# install Gramps addons
RUN mkdir -p /root/gramps/gramps$GRAMPS_VERSION/plugins && \
    wget https://github.com/gramps-project/addons/archive/refs/heads/master.zip && \
    for addon in PostgreSQL SharedPostgreSQL FilterRules JSON; do \
        unzip -p master.zip addons-master/gramps$GRAMPS_VERSION/download/$addon.addon.tgz | \
        tar -xvz -C /root/gramps/gramps$GRAMPS_VERSION/plugins; \
    done && \
    rm master.zip