RICardo Project, Historical Trade Database
This project contains 2 applications:
ricardo_api': a Flask application exposing a REST API serving data from the sqlite database in JSON format.client: an HTML5 angular.js application displaying data vizualisations from the ricardo API.
The corresponding data-set is hosted at https://github.com/medialab/ricardo_data
If you want to run your instance of RICardo locally on your machine, be sure you have the following requirements installed.
Clone RICardo from the command line:
$ git clone https://github.com/medialab/ricardo.git
Browse to RICardo root folder:
$ cd ricardo
Browse to RICardo client root folder:
$ cd ricardo/client
Install client-side dependencies:
$ npm install
Edit configuration file
$ cd src/js
$ cp config.sample.js config.js
$ vi config.js
# edit config.js and add the correct root to API
'use strict';
angular.module('ricardo')
.constant('BASE_API_URL', 'http://localhost:5000')
You can now run RICardo with the following command :
$ npm start
Once this is running, go to http://localhost:8080/
You can build the RICardo application with the following command :
$ npm run build
The build will be available in the dist folder.
Controllers are located under the src/js/controllers folder with one file per controller, and you should reference them in the file src/js/controllers/index.js.
Directives are located under the src/js/directives folder with one file per directive, and you should reference them in the file src/js/directives/index.js. Every chart component should have a directive.
Angular partials are located under the folder src/public/partials.
Api calls are done in the service src/js/services.js.
All files under the src/public folder are static.
CSS files must be referenced in the file src/js/style.js. The build process generates one CSS file from this JS file, and include it (with a style tag) in the index.html.
The same philosophy is applied to js dependencies, with the file src/js/external_dependencies.
If you want to run the RICardo API locally on your machine, be sure you have the following requirements installed.
The dataset needed to run this application is available in the ricardo data repository.
You can follow the readme of the project to know how to build the database.
It should produce a file called RICardo_viz.sqlite
You must copy RICardo_viz.sqlite in the folderapi/ricardo_api.
You can change the path of the database by editing the api/ricardo_api/config.py if needed.
Go into the Api folder of the project
cd api
Install the server dependencies
pip install -r requirements.txt
Launch the server
python runserver.py
+api/
|__+ricardo_api
|__ __init.py__ -> connection to database
|__config.py -> information element of database
|__models.py -> methods to get and transform datas from DB to JSON api responses
|__views.py -> routes to serve datas
|__RICardo_visualisation.sqlite
# RICentities
List all entities from the database
# reporting_entities
List entities which has reported Imp/Exp flows.
# flows
The main raw data API. flows will provide the exp/imp flows between countries.
# continent_flows
Flows aggregated by continent
# flows_sources
Provide a list of all sources used in a flow API call.
Used to get them all at once without repeating them in flow.
At the root of this project, there is a folder docker that contains a https://docs.docker.com/compose/ stack
for the development.
You can run it by following those steps :
cd docker
docker-compose up
It will start for you the project stack with :
- An nginx that serves the client on port 80
- the API
In DEV mode (check the file .env), the auto-reload of the code is enabled.
The docker images of the project (ie. API & client) are built automatically by a gitlab process (see .gitlab-ci.yml).
Those images are built with the Dockerfiles client/Dockerfile & api/Dockerfile.
To test those images on local before pushing to prod you can use the dedicated docker-compose which set the data version to use.
cd docker
docker-compose -f docker-compose.test-prod.yml -p ricardo-test-prod up --build