Skip to main content Link Menu Expand (external link) Document Search Copy Copied
Powerful Personal Data

Backend

Description

The backend server’s goal is to convert data in non linked format to JSON-LD or turtle representations. This is achieved by using Rml.io which allows the user to write RML rules that can automatically convert documents to RDF. Currently the backend supports the conversion of weather data from OpenWeatherMap to JSON-LD.

The endpoint can be accessed at localhost:3000/weather/$cityName$

The backend server will retrieve the longitude and latitude of the city name using the OpenStreetMap API which then get passed to the weather endpoint. The result is converted to RDF using a JavaScript wrapper over RmlMapper. The latest version of the JAR file for the RmlMapper can be downloaded using the following command:

node ./node_modules/@rmlio/rmlmapper-java-wrapper/lib/download-latest-rmlmapper.js.

The RML rules used to convert the data can be found at /mappings/final_RML.ttl. This specification has been generated using Yarrrml which allowed us to use YAML to easily generate the rules, and then convert them to turtle.

Dependencies

  • The backend server runs using Node.js 18 (needed for the built-in fetch functionality) and Fastify. This allows us to have a lightweight API.
  • The conversion to RDF is done using RmlMapper which is built in Java, and as a consequence we need to have JDK18 as a dependency for running the application.

Local installation

To install and run the server locally, follow the steps below:

cp .env .env.example # fill in the environment variables
node ./node_modules/@rmlio/rmlmapper-java-wrapper/lib/download-latest-rmlmapper.js # Download the latest version of the JAR file
npm i
npm run start

Deploy

The server is currently deployed using DigitalOpen App Platform and can be accessed at: https://idlab.osoc.be/weather/Brussels.

The same docker image used to deploy the server can be used locally too. To do that, follow the instructions below:

docker build . -t powerful-personal-data
docker run -p 127.0.0.1:3000:3000 powerful-personal-data # you can now access the service by going to http://127.0.0.1:3000