diff --git a/docker/Dockerfile b/docker/Dockerfile --- a/docker/Dockerfile +++ b/docker/Dockerfile @@ -1,16 +1,30 @@ FROM adoptopenjdk/openjdk11:debian-jre +RUN export DEBIAN_FRONTEND=noninteractive && \ + apt-get update && apt-get upgrade -y && \ + apt-get install -y python3 python3-click python3-requests wget && \ + apt-get clean && \ + rm -rf /var/lib/apt/lists/* + # Download and install jars -ADD https://github.com/javasoze/clue/releases/download/release-6.2.0-1.0.0/clue-6.2.0-1.0.0.jar /opt/ -ADD https://repo1.maven.org/maven2/org/apache/maven/indexer/indexer-cli/6.0.0/indexer-cli-6.0.0.jar /opt/ +RUN cd /opt && wget -q \ + https://github.com/javasoze/clue/releases/download/release-6.2.0-1.0.0/clue-6.2.0-1.0.0.jar +# coming out of https://github.com/javasoze/clue/releases/tag/release-6.2.0-1.0.0 +RUN cd /opt && wget -q \ + https://repo1.maven.org/maven2/org/apache/maven/indexer/indexer-cli/6.0.0/indexer-cli-6.0.0.jar +# FIXME: Retrieve https://repo1.maven.org/maven2/org/apache/maven/indexer/indexer-cli/6.0.0/indexer-cli-6.0.0.jar.sha1 +# indexer-cli-6.0.0.jar.sha1: eeb98596b7fed4aa13fa13ecafcbb843ef8ab697 # Copy index extraction script + +COPY run_full_export.py /opt/ COPY extract_indexes.sh /opt/ +RUN mkdir /work/ && chmod +x /opt/extract_indexes.sh WORKDIR /work/ -RUN ls /opt/ -RUN ls -R /work/ +ENV MVN_IDX_EXPORTER_BASE_URL=https://repo.maven.apache.org/maven2/ +ENV MVN_IDX_EXPORTER_WORK_DIR=/work +ENV MVN_IDX_EXPORTER_PUBLISH_DIR=/publish -# Parse default index file (will be overriden by cli parameters) -CMD ["sh", "/opt/extract_indexes.sh"] +CMD ["python3", "/opt/run_full_export.py"] diff --git a/docker/extract_indexes.sh b/docker/extract_indexes.sh --- a/docker/extract_indexes.sh +++ b/docker/extract_indexes.sh @@ -5,7 +5,7 @@ # License: GNU General Public License version 3, or any later version # See top-level LICENSE file for more information -WORKDIR=/work +WORKDIR=${1-/work} FILE_IN=$WORKDIR/nexus-maven-repository-index.gz localtime=$(date +"%Y-%m-%d %H:%M:%S") diff --git a/scripts/run_full_export.py b/docker/run_full_export.py rename from scripts/run_full_export.py rename to docker/run_full_export.py --- a/scripts/run_full_export.py +++ b/docker/run_full_export.py @@ -9,6 +9,7 @@ import datetime import glob import re +from subprocess import check_call from os import chdir, makedirs from os.path import getsize, isabs, isdir, isfile, join, basename from pathlib import Path @@ -17,10 +18,6 @@ import requests -import docker -from docker.client import DockerClient -from docker.models.images import Image - logger = logging.getLogger(__name__) @@ -28,51 +25,10 @@ MAVEN_INDEX_ARCHIVE = f"{MAVEN_INDEX_NAME}.gz" -def _docker_image_get( - client: DockerClient, docker_image_name: str, docker_image_update: bool = False -) -> Image: - """Retrieve docker image locally.""" - if docker_image_update: - return client.images.pull(repository=docker_image_name) - try: - image = next(iter(client.images.list(name=docker_image_name))) - except StopIteration: - logger.info("Docker: Could not find %s. Pulling it.", docker_image_name) - image = client.images.pull(repository=docker_image_name) - else: - logger.info( - "Docker: Found image %s locally, ID is %s.", image, image.attrs["Id"] - ) - return image - - -def _extract_indexes( - work_dir: str, docker_image_name: str, docker_image_update: bool = False -) -> None: - """Extract indexes out of the working directory. - - As an implementation details, this starts the maven indexer exporter container using - a docker image. This will use the local image if present, otherwise pull it from - docker hub first. - - """ - # Initialise the docker client. - client = docker.from_env() - image = _docker_image_get(client, docker_image_name, docker_image_update) - # Run the extraction process through the docker image (which runs the extract index - # script), see ../docker/Dockerfile. - ret = client.containers.run( - image, - tty=True, - volumes={work_dir: {"bind": "/work", "mode": "rw"}}, - ) - - logger.info("Docker log:\n%s", ret.decode()) - - def _download_indexes(work_dir: str, instance_url: str) -> None: - """ Download all required indexes from the .index/ directory + """Download all required indexes from the .index/ directory of the specified instance. + """ logger.info("Downloading all required indexes") @@ -143,16 +99,7 @@ help="Absolute path to the final directory.", default="/tmp/maven-index-exporter/publish/", ) -@click.option( - "--docker-image-name", help="Docker image", default="maven-index-exporter" -) -@click.option( - "--docker-image-update", - is_flag=True, - help="Trigger a docker image update.", - default=False, -) -def main(base_url, work_dir, publish_dir, docker_image_name, docker_image_update): +def main(base_url, work_dir, publish_dir): now = datetime.datetime.now() logger.info("Script: run_full_export") logger.info("Timestamp: %s", now.strftime("%Y-%m-%d %H:%M:%S")) @@ -177,10 +124,8 @@ # Only fetch the new ones, existing files won't be re-downloaded. _download_indexes(work_dir, base_url) - # Run Docker on the downloaded indexes. - _extract_indexes( - work_dir, docker_image_name, docker_image_update=docker_image_update - ) + # Extract indexes into a .fld file + check_call(["/opt/extract_indexes.sh", work_dir]) logger.info("Export directory has the following files:") export_dir = join(work_dir, "export") @@ -218,4 +163,4 @@ if __name__ == "__main__": logging.basicConfig(level=logging.INFO) - main() + main(auto_envvar_prefix='MVN_IDX_EXPORTER')