datahub/docker/datahub-ingestion/Dockerfile

# Defining environment
ARG APP_ENV=full
ARG BASE_IMAGE=acryldata/datahub-ingestion-base
ARG DOCKER_VERSION=head-full
ARG DEBIAN_REPO_URL=https://deb.debian.org/debian
ARG UBUNTU_REPO_URL=http://ports.ubuntu.com/ubuntu-ports
ARG PIP_MIRROR_URL=https://pypi.python.org/simple

FROM $BASE_IMAGE:$DOCKER_VERSION AS base

# Optionally set corporate mirror for deb
USER 0
ARG DEBIAN_REPO_URL
ARG UBUNTU_REPO_URL
RUN if [ "${DEBIAN_REPO_URL}" != "http://deb.debian.org/debian" ] ; then sed -i "s#http.*://deb.debian.org/debian#${DEBIAN_REPO_URL}#g" /etc/apt/sources.list ; fi
RUN if [ "${UBUNTU_REPO_URL}" != "http://ports.ubuntu.com/ubuntu-ports" ] ; then sed -i "s#http.*://ports.ubuntu.com/ubuntu-ports#${UBUNTU_REPO_URL}#g" /etc/apt/sources.list ; fi
USER datahub

# Optionally set corporate mirror for pip
ARG PIP_MIRROR_URL
RUN if [ "${PIP_MIRROR_URL}" != "https://pypi.python.org/simple" ] ; then pip config set global.index-url ${PIP_MIRROR_URL} ; fi
ENV UV_INDEX_URL=${PIP_MIRROR_URL}

COPY --chown=datahub ./metadata-ingestion /metadata-ingestion
COPY --chown=datahub ./metadata-ingestion-modules/airflow-plugin /metadata-ingestion/airflow-plugin

ARG RELEASE_VERSION
WORKDIR /metadata-ingestion
RUN sed -i.bak "s/__version__ = \"1\!0.0.0.dev0\"/__version__ = \"$(echo $RELEASE_VERSION|sed s/-/+/)\"/" src/datahub/__init__.py && \
    sed -i.bak "s/__version__ = \"1\!0.0.0.dev0\"/__version__ = \"$(echo $RELEASE_VERSION|sed s/-/+/)\"/" airflow-plugin/src/datahub_airflow_plugin/__init__.py && \
    cat src/datahub/__init__.py | grep __version__ && \
    cat airflow-plugin/src/datahub_airflow_plugin/__init__.py | grep __version__

FROM base AS slim-install

RUN --mount=type=cache,target=/datahub-ingestion/.cache/uv,uid=1000,gid=1000 \
    UV_LINK_MODE=copy uv pip install -e ".[base,datahub-rest,datahub-kafka,snowflake,bigquery,redshift,mysql,postgres,hive,clickhouse,glue,dbt,looker,lookml,tableau,powerbi,superset,datahub-business-glossary]"

FROM base AS full-install-build

USER 0
RUN apt-get update && apt-get install -y -qq maven

USER datahub
COPY ./docker/datahub-ingestion/pyspark_jars.sh .

RUN --mount=type=cache,target=/datahub-ingestion/.cache/uv,uid=1000,gid=1000 \
    UV_LINK_MODE=copy uv pip install -e ".[base,all]" "./airflow-plugin[plugin-v2]" && \
    ./pyspark_jars.sh && \
    datahub --version

FROM base AS full-install

COPY --from=full-install-build ${VIRTUAL_ENV} ${VIRTUAL_ENV}

FROM base AS dev-install
# Dummy stage for development. Assumes code is built on your machine and mounted to this image.
# See this excellent thread https://github.com/docker/cli/issues/1134

FROM ${APP_ENV}-install AS final

WORKDIR /datahub-ingestion

USER datahub