Mapping predictor backend

This project is part of my work done during my Google Summer of Code 2018 in collaboration with DBpedia

This project contains all the API calls that can be done with from the API and some others only accesible with an external HTTP client, like Insomnia.

API documentation

The API has been documented using Sphynx. Can be consulted on the /docs folder on this repo or online on Github Pages

Develop environment

This is a Java EE project, so it is recommended to use an advanced IDE, as Intellij IDEA. To run the application, as a Java EE implementation is required, I recommend to use Glassfish 5.0.1 or higher.

Docker environment

To ease the development process, as well as the deploying step, I've containerized this application with Docker.

This container depends from a SQL database. I recommend to use MySQL. It supports version 8 or higher. Also, other MySQL implementations like MariaDB have been also tested successfully.

Docker compose is used to deploy the full web application, including both backend and frontend containers.

$ docker-compose up -d

To build the standalone docker image:

$ docker build -t vfrico/dbpedia-mappings-backend:latest .

The image is hosted on the docker registry as: vfrico/dbpedia-mappings-backend

Installation of the backend

In order to the backend to work properly, it is needed to install on the database the tables that the system is going to need. Is also possible to load a set of sample annotations.

Both tasks can be done after the backend container is ready on this URI. Click first on Create tables button and after a few seconds, on add CSV.

http://ip:8080/predictor

Alternatively, you can use an HTTP client to submit the following requests:

POST /installation/createtables
POST /installation/addAllCSV

HTTP Client

It is recommended to use an HTTP client like Insomnia. It is licensed under the GPLv3 license and has a very handful way to manage environment variables and management of HTTP requests by organizing them on folders.