Skip to content

Latest commit

 

History

History
165 lines (114 loc) · 5.08 KB

README.md

File metadata and controls

165 lines (114 loc) · 5.08 KB
Raw

Spatial Data Package Platform

A web application for publishing and collaborating around urban planning projects. This project is used to run Gemeindescan.ch, a portal for holistic urban assessments and tool for setting planning priorities ("Gemeinden" are municipalities in German). This project is being developed by cividi GmbH, and supported as an Engagement Migros pioneer initiative.

For more information, please visit Gemeindescan.ch and read our Whitepaper.

Additional documentation

Developer guide

🚧 The current tech stack of this project is:

  • Vuetify
  • Vue.js
  • GraphQL
  • Graphene
  • Django
  • Wagtail
  • Docker

For a tour of the app and a testing protocol see docs/testing.md.

Setup project

(1) Prepare local configuration

  • Add www.local and django to the 127.0.0.1 entry in /etc/hosts (for screenshot service).
  • A symlink to the nginx configuration file in the www container will be created automatically (/etc/nginx/conf.d/www.local.conf) for production use a dedicated copy
  • for production or custom settings add an .env file with the following environment variables:
DJANGO_SECRET=some.R4nd0m_k3y
MAPBOX_TOKEN=a.mapbox.api.token
MAPBOX_STYLE=mapbox://styles/gemeindescan/ck6rp249516tg1iqkmt48o4pz

...which will be sourced by the Makefile scripts. Or use some other way to configure your environment.

If you want to add support for Fathom analytics add these two optional environment variables as well:

FATHOM_SITEID=you.fathom.siteid
FATHOM_URL=custom.domain.net # needs extra setup

(2) Download containers and start them

docker-compose pull
make up
make init

(3) If make init doesn't work you can execute the commands one by one.

# go into django container
make enter_django
cd django
python3 manage.py migrate
python3 manage.py createsuperuser
make import-gemeinden-json
# exit django container
exit

For development we use a named docker volume for the `var` folder inside the container, this named
volume will be created automatically. This makes setup under windows easier.
For production under linux a bind mount is used, for easier backups and no accidental data deletion.

Database files and static generated files are stored in `var` folder.

### Start containers

```bash
make up

The front- and backend should now be accessible via

Frontend: http://www:8000
Backend:  http://www:8000/gmanage
GraphQLi: http://www:8000/graphql

start services

Start all the services at the same time using

make start_all

or individually using:

# 1. shell
make start_vue
# 2. shell
make start_django
# 3. shell
make start_screenshotservice

You can also start the vue / django service from the vscode shell (see editor).

stop services

make stop

Unit testing

make tests

This command runs first the backend tests and afterwards the frontend test.

For the backend tests a new test database is generated and the fixtures are loaded.

IDE notes

We use Visual Studio Code (vscode) as editor. Any other editor is also fine. Here are some notes to tweak your IDE.

vsvode

With vscode the plugins are automatically installed and linting and autocomplete works out of the box, with the data from inside the container.

You need to have installed the vscode extension Remote - Containers in your local vscode installation.

You need to start two editors with code vue and code django (install code command in path from vscode). After they started you need to click on reopen in container (first time start takes a bit longer to install plugins).

vue vscode

Type make in the vscode terminal, that creates the node dev server on your maschine which is accessible via http://www:8000. Make sure to add www and www.local to your /etc/hosts file to point to 127.0.0.1.

django vscode

Type make in the vscode terminal, that creates the django dev server on your maschine on the url http://www:8000/gmanage.

Push images to docker hub

  • update the version number in the Makefiles in the docker folders and docker-compose.yml
  • make build for all images in the root, or in the django or vue subfolders, this builds the docker images
  • make push for all images in the root, or in the django or vue subfolders, this uploads the docker images to docker hub

Deploy to production

Type make deploy_prod in the project root. The file env.hosts.prod needs to have the right settings.

You may also wish to copy docker-compose.yml and tweak to other hosting needs, and start your containers by specifying the alternative configuration:

docker-compose -f docker-compose.prod.yml up