Skip to content

Latest commit

 

History

History
156 lines (98 loc) · 5.52 KB

setup-dev.md

File metadata and controls

156 lines (98 loc) · 5.52 KB
Raw

Setup a dev environment

Using Docker

Docker is the easiest way to install the Taxonomy Editor, play with it, and even modify the code.

Docker provides an isolated environment, very close to a Virtual Machine. This environment contains everything required to launch the Taxonomy Editor. There is no need to install any modules separately.

Prerequisites

If you run e.g. Debian, don't forget to add your user to the docker group!

Setup

The docker-compose for Taxonomy Editor is designed in such a way that all development servers can be built rapidly.

After first checkout, or on requirements or important element changes, run the following commands in the root directory:

DOCKER_BUILDKIT=1 docker-compose build

And every time, to get the server running, just use:

docker-compose up

You should have a running environment.

The React application is exposed at: http://ui.taxonomy.localhost:8091

The API is exposed at: http://api.taxonomy.localhost:8091

You can also access the Neo4j Admin Console at http://localhost:7474/browser/

If you modify any file in the React App, the changes will be taken into account instantly. However, this feature is not compatible with Windows systems. In order to use live reload on a Windows machine, you will need to install and use WSL2. This will allow you to run the development server in a Linux environment, and the live reload feature will work as expected.

If you modify any files related to the Python API, you need to restart the taxonomy_api container in Docker: docker-compose restart taxonomy_api

You can modify the .env file to fit your needs, but you should not commit any changes that are not defaults to the project. Notably, if you use a uid which is not 1000, you should personalize the USER_UID variable.

A smarter way to customize things, is to use direnv with a .envrc file, or to simply have a script to source (see the . command) to load environment variables (they have priority above .env).

Creating Pull Requests with the Taxonomy Editor

If you are using docker, the .env file must be updated in order to use the feature. If you'd like to see it in action in development mode, do the following steps:

  • Make sure a fork of openfoodfacts-server repository is created in your Github account.
  • Update the access_token variable in .env with your personal access token.
  • Update the repo_uri variable in .env with your Github for uri: (username/repository_name)

That's it! Now, you'll be able to view any created PR's in your fork of Taxonomy Editor.

Importing some test data

docker-compose run --rm taxonomy_api sample/load.py sample/test-neo4j.json

Add the --reset switch if you already have data you want to remove.

Going into the Neo4j Admin Console, you should be able to see the data.

Without Docker

Install Neo4j

You are required to install Neo4j in your system.

Neo4j Desktop (proprietary) is an option, or get the open-source community version.

Visit this link for more information.

Launch it and verify it's working by visiting the Neo4j browser: http://localhost:7474/browser/. In the admin UI, you should connect to localhost:7687.

Install the backend in a Python Virtualenv

Go to the backend directory of the Taxonomy Editor.

Create a virtualenv with python:

python3 -m venv .venv

and activate it:

source .venv

Install dependencies:

source pip install -r requirements.txt

Start the server locally:

uvicorn editor.api:app --host 127.0.0.1 --port 8080

Open http://127.0.0.1:8080 in your browser to check out the API.

You will see the following:

{"message": "Hello user! Tip: open /docs or /redoc for documentation"}

Install the npm dev server

(Note: this might not be needed if you only want to work on the backend)

You must have NodeJS and npm installed on your system. (LTS - 16.17.0, currently). See https://nodejs.org/en/download/ and https://docs.npmjs.com/cli/v7/configuring-npm/install for further installation steps.

Go to the taxonomy-editor-frontend/ directory of the Taxonomy Editor.

Then run:

npm install .

Run the server with:

REACT_APP_API_URL="http://localhost:8080/" npm start

(or)

export REACT_APP_API_URL="http://localhost:8080/"
npm start

This should automatically open the Taxonomy Editor frontend in this URL: http://localhost:8080/

Importing some test data

To import some test data, in your Python virtualenv (don't forget to activate it), in the backend directory:

python3 sample/load.py sample/test-neo4j.json

Add the --reset switch if you already have data you want to remove.

Going into the Neo4j Admin Console, you should be able to see the data.