API for reading from the Content Graph/Store. Currently, the API is capable of interfacing with two databases:
Stardog (may be generalised to any SPARQL1.1 store), and Elasticsearch. Client specific code resides in the
clients directory and the clients are listed in api.py, all other code is/should be client-agnostic.
This fork has been modified to run a development setup using Docker Compose.
You should be able to:
docker-compose pull
docker-compose build
docker-compose up
A Travis CI job for this repository can be viewed at:
The above Travis job pushes the built Docker image to Docker hub here:
NB the Dockerfile mentioned below has been renamed to Dockerfile.bbc.
python3 -m venv env
. env/bin/activate
pip3 install -r requirements.txt
export DB_CLIENT='stardog'
export DB_ENDPOINT=http://$SERVER:$PORT/content-graph-test/query
export DB_USER=$USERNAME
export DB_PASS=$PASSWORD
export DB_CLIENT='elasticsearch'
export DB_ENDPOINT=http://localhost:9200
export DB_USER=$USERNAME
export DB_PASS=$PASSWORD
Resources to run Elasticseaerch stack locally.
PORT=5001 \
PYTHONPATH=.:$PYTHONPATH \
python -m app.api
4. Visit the application at http://localhost:5001.
Check if your code is PEP8 compliant:
pycodestyle app --max-line-length=119
Run tests using:
PYTHONPATH=.:$PYTHONPATH py.test --cov-report term-missing --cov=app tests/ --cov-branch -vv
NB: SPARQL client not fully covered - still need to refactor tests.
This app is provided as with a Dockerfile which is used to build a container. This should then be pushed to a container registry and deployed either as a manual process or using something such as build triggers and a continuous delivery platform like Spinnaker.