Table of Content

• Project Info
• Repository Structure
• Docker Setup
• Local Setup
  • Start the Database
    • Database: Normal Start
    • Database: Initial Start / Reset
  • Generate Data
  • Start the Full Text Search
    • Run Frontend with Docker
    • Run Frontend locally
• Further Documentation
• Notes

Project Info

The platform is our contribution to democratizing access to political debates and issues.

Open Discourse is a non-profit project of the employees of Limebit GmbH. The idea emerged from the skills and motivations of the employees, in break conversations and from the common ideas of democracy.

We hope that through our preliminary work, data-based journalism, science and civil society will benefit and that the facilitated access to data will encourage to analyze the political history of the Bundestag based on the language used by politicians.

Repository Structure

This Repo is structured in three different parts.

• database:
  • Docker-Container for the Postgres Database
  • Contains Scripts that update the Database
• frontend:
  • Frontend for the Full Text Search
• proxy:
  • Docker-Container for the Proxy, which protects the database
• python:
  • Includes every python script in different subsections, sorted by execution order

Docker Setup

For a quick setup using Docker, please read the DOCKER_SETUP

Local Setup

Required software: python3, yarn, docker-compose, node version 12 - ideally installed via node version manager (nvm)

• run yarn in following directories:
  • database
  • frontend
• run sh setup.sh in the python directory
• run docker-compose build in the root folder

Start the Database

These steps will guide you through starting the Database

Database: Normal Start

You can easily start the Database via docker-compose.

// run from repository root
docker-compose up -d database

Database: Initial Start / Reset

For the initial start of the Database, you will also need to upload the schema.

// run from database folder
yarn run db:update:local

Generate Data

Generate the OpenDiscourse-Database from the ground up. The Database has to be started for this script to finish.

This script is just a pipeline executing all scripts in src. You can also manually run every script seperatly. For Documentation on this, please visit the README in src

// run from python folder
sh build.sh

Start the Full Text Search

Note: All of the previous steps have to be completed at least once for the Full Text Search to work properly.

If you want to setup the Full Text Search, follow these steps:

• run yarn in following directories:
  • frontend
  • proxy

Choose one of the following ways to start the Frontend:

Run Frontend with Docker

• run docker-compose up -d in the root folder

Run Frontend locally

• run docker-compose up -d database proxyin the root folder
• run yarn dev in the frontend folder

Further Documentation

• Documentation of the database can be found in the README in database
• Documentation of the frontend can be found in the README in frontend
• Documentation of the proxy can be found in the README in proxy
• Documentation of the python service can be found in the README in python
• Documentation of every python-script can be found in the README in python/src

Notes

• We use Python 3.7.4 during development of the project
• The graphql endpoint was deprecated and removed by version 1.1.0