docker-wagtail-develop

A script to painlessly set up a Docker environment for development of Wagtail - inspired by vagrant-wagtail-develop

Initial work in Bristol sprint January 2020 by esperk and saevarom.

Setup

Requirements: Docker and Docker Compose (Docker Compose is included with Docker Desktop for Mac and Windows).

Open a terminal and follow those instructions:

# 1. Decide where to put the project. We use "~/Development" in our examples.
cd ~/Development
# 2. Clone the docker-wagtail-develop repository in a new "wagtail-dev" folder.
git clone https://github.com/wagtail/docker-wagtail-develop.git wagtail-dev
# 3. Move inside the new folder.
cd wagtail-dev/
# 4. Run the setup script. This will check out the bakerydemo project and local copies of wagtail and its dependencies.
./setup.sh
# 5. Build the containers
docker compose build

It can take a while (typically 15-20 minutes) to fetch and build all dependencies and containers.

Here is the resulting folder structure:

.
├── libs          # Supporting libraries to develop Wagtail against.
├── wagtail       # Wagtail repository / codebase.
└── bakerydemo    # Wagtail Bakery project used for development.

Once the build is complete:

# 6. Start your containers and wait for them to finish their startup scripts.
docker compose up web

You might see a message like this the first time you run your containers. This is normal because the frontend container has not finished building the assets for the Wagtail admin. Just wait a few seconds for the frontend container to finish building (you should see a message like webpack compiled successfully in 15557 ms and then stop and start your containers again (Ctrl+C + docker compose up).

WARNINGS:
?: (wagtailadmin.W001) CSS for the Wagtail admin is missing
	HINT:
            Most likely you are running a development (non-packaged) copy of
            Wagtail and have not built the static assets -
            see https://docs.wagtail.org/en/latest/contributing/developing.html

            File not found: /code/wagtail/wagtail/admin/static/wagtailadmin/css/normalize.css```

# 7. Now in a new shell, run the database setup script. The database will be persisted across container executions by Docker's Volumes system so you will only need to run this command the first time you start the database.
./setup-db.sh
# Success!

If you're running this on Linux you might get into some privilege issues that can be solved using this command (tested on Ubuntu):

CURRENT_UID=$(id -u):$(id -g) docker compose -f compose.yaml -f compose.linux.yaml up

Alternatively, if you're using VSCode and have the "Remote - Containers" extension, you can open the command palette and select "Remote Containers - Reopen in Container" to attach VSCode to the container. This allows for much deeper debugging.

• Visit your site at http://localhost:8000
• The admin interface is at http://localhost:8000/admin/ - log in with admin / changeme.

What you can do

See a list of running containers

$ docker compose ps
  Name                Command               State           Ports
--------------------------------------------------------------------------
db         docker-entrypoint.sh postgres    Up          5432/tcp
frontend   docker-entrypoint.sh /bin/ ...   Up
web        ./manage.py runserver 0.0. ...   Up          0.0.0.0:8000->8000/tcp

Build the backend Docker image

make build

or

docker compose build web

Bring the backend Docker container up

make start

or

docker compose up

Stop all Docker containers

make stop

or

docker compose stop

Stop all and remove all Docker containers

make down

or

docker compose down

Run tests

make test

or

docker compose exec -w /code/wagtail web python runtests.py

Run tests for a specific file

make test file=wagtail.admin.tests.test_name

or

docker compose exec -w /code/wagtail web python runtests.py wagtail.admin.tests.{test_file_name_here}

Format Wagtail codebase

make format-wagtail

or

docker compose exec -w /code/wagtail web make format-server
docker compose exec frontend make format-client

Lint Wagtail codebase

make lint-wagtail

or

docker compose exec -w /code/wagtail web make lint-server
docker compose exec -w /code/wagtail web make lint-docs
docker compose exec frontend make lint-client

Open a Django shell session

make ssh-shell

or

docker compose exec web python manage.py shell

Open a PostgreSQL shell session

make ssh-db

or

docker compose exec web python manage.py dbshell

Open a shell on the web server

make ssh

or

docker compose exec web bash

Open a shell to work with the frontend code (Node/NPM)

make ssh-fe

or

docker compose exec frontend bash

Open a shell to work within the wagtail container

make ssh-fe

or

docker compose exec -w /code/wagtail web bash

Make migrations to the wagtail bakery site

make migrations

or

docker compose exec web python manage.py makemigrations

Migrate the wagtail bakery site

make migrate

or

docker compose exec web python manage.py migrate

Destroy the current database volume

If this project has changed database versions since you first started using it, you may need to delete the database volume and then recreate the bakerydemo data using the setup-db.sh script. To delete all the data from the database running in the db container, run the following commands:

make down
make destroy-db

or

docker compose down
docker volume rm docker-wagtail-develop_postgres-data

Getting ready to contribute

Here are other actions you will likely need to do to make your first contribution to Wagtail.

Set up git remotes to Wagtail forks (run these lines outside of the Docker instances, on your machine):

cd ~/Development/wagtail-dev/wagtail
# Change the default origin remote to point to your fork.
git remote set-url origin git@github.com:<USERNAME>/wagtail.git
# Add wagtail/wagtail as the "upstream" remote.
git remote add upstream git@github.com:wagtail/wagtail.git
# Pull latest changes from all remotes / forks.
git pull --all

Contributing to Willow

You can use the same setup to contribute to Willow. You simply do the same operations to fork the Willow project and point your local copy of Willow to your fork.

See also

• Vagrant Wagtail development

Testing with Elasticsearch

To test the Wagtail search with the Elasticsearch backend or OpenSearch, use the additional service:

docker compose up web elastic

Then install any dependencies and configure Wagtail per the developer documentation, configuring the search backend to use http://elastic:9200 as the URL.