The platform that powers republik.ch. A tailored solution for a membership based online magazine.
If you are a Republik reader (thank you!) and you have any issues or feature requests, please do kindly direct them to our support team. Unfortunately we can also not currently accept external contributions to our codebase, as we are a small team and want to keep our focus on improving and maintaining the product for all of our readers. Thank you for your interest in case you were considering it, we are impressed.
This turborepo uses Yarn as a package manager. It includes the following apps and packages:
www: providing the public frontendpublikator: providing publication management interfaceadmin: providing the customer management interfaceapi: providing the graphql apiassets: fetching, compressing and resizing assets
The frontends are Next.js apps, the backends use Express.js.
styleguide: a React component library shared by all frontends and used by theapito render newslettersbackend-modules/*: packages used by theapiandassetsservermdast/*: packages used to work with mdast (previously in it's own monorepo)eslint-config-*: multiple eslint configurations which are used accross our appsnextjs-apollo-client: a reusable Nextjs / Apollo Client setuptsconfig: a package to store shared typescript configurationsicons: A package to make SVG-files available as react.js components to be used across our frontends.
All packages and apps support TypeScript and plain ECMAScript.
The logo and fonts are the property of their owners (logo—Project R, GT America—GrilliType and Rubis—Nootype), and may not be reproduced without permission.
The www, publikator app and styleguide are BSD-3-Clause licensed. The api, assets app and all backend-modules are AGPL-3.0 licensed. See respective license files in subfolders.
This turborepo has some additional tools already setup for you:
To get started you'll need:
- yarn v1.22
- Node.js v20
- Docker or native postgresql@14, elasticsearch@6 and redis
Setup with Docker
The included docker-compose.yml starts all external-services. Currently that's: postgresql, redis, elasticsearch (and kibana).
The data is persisted in ./docker-data/.
docker-compose up [-d]
We recommend you install the postgresql client tools on your machine to interact with the database. The tests scripts also depend on the clients being installed.
# linux
sudo apt install postgresql-client-12
When postgresql is running in docker, client tools like psql or createdb/dropdb don't automatically connect to it. They try to access postgresql via a local socket, when instead you want them to connect via network to localhost. To make your life easier, you can add the following environment variables to ~/.bashrc / ~/.zshrc so the client tools connect to localhost per default.
export PGHOST=127.0.0.1
export PGUSER=postgres
Native Setup with Homebrew
brew install postgresql@12 elasticsearch@6 redis nvm
nvm install 14
nvm alias default 14
npm install -g [email protected]
brew services start postgresql@12
brew services start elasticsearch@6
brew services start redisdocker run -p 5601:5601 -e ELASTICSEARCH_HOSTS=http://host.docker.internal:9200 docker.elastic.co/kibana/kibana-oss:6.7.0Note:
- Elasticsearch and Kibana versions must match, ckeck ES version at
http://localhost:9200/ ELASTICSEARCH_HOSTSmust be accessible within docker.
All apps and the styleguide provide an .env.example, the provided default values should be enough to get started:
cp apps/www/.env.example apps/www/.env
cp apps/publikator/.env.example apps/publikator/.env
cp apps/admin/.env.example apps/admin/.env
cp apps/api/.env.example apps/api/.env
cp apps/assets/.env.example apps/assets/.envMigrating from Individual Repos
You may copyover your environment from the individual repos with one manual edit:
cp ../republik-frontend/.env apps/www/.env
cp ../publikator-frontend/.env apps/publikator/.env
cp ../republik-admin-frontend/.env apps/admin/.env
cp ../styleguide/.env packages/styleguide/.env
cp ../backends/.env apps/api/.env
echo "PORT=5010" >> apps/api/.env
cp ../backends/servers/assets/.env apps/assets/.envThe COOKIE_NAME env-variable can be defined in apps/api and apps/www. It's crucial that the value defined in the apps/www file matches the one of the API that you're developing against.
If these two env-variables don't match, www will be stuck in a redirection loop when trying to open the URL /.
In addition to the cookie name, there the following env-variables must be set to allow for token based authentication.
The env-variables JWT_COOKIE_NAME and JWT_ISSUER are present in both the api and www env-files and must be identical.
Additionally a private-key must be provided to the api with theJWT_PRIVATE_KEY env-variable and a public-key must be provided to www with the JWT_PUBLIC_KEY env-variable.
There is a script under scripts/generate-keypair.sh that can generate keys in the right format to be directly passed into the corresponding env-files.
yarn install
yarn build
yarn dev:setupTo develop all apps and packages, run the following command:
yarn devPlease be patient on boot. It might take a minute for everything to compile and a few nodemon restarts before everything runs smoothly.
If you don't want all apps to run when using the dev script, you can use the filter flag.
(see the Turborepo documentation)
For example if you only want to run the republik frontend run yarn dev --filter=@orbiting/www-app.
In most cases you have certain dependencies that should be run as well, for example the styleguide if you're developing in the frontend. In that case simply append ... directly after the filter, to ensure that the additionally to the filtered app, all dependencies are executed as well. (For example in www run: yarn dev --filter=@orbiting/www-app...)
<type>(<scope>): <subject>
<BLANK LINE>
<body>
<BLANK LINE>
<footer>
You can use yarn commit to generate a message via an interactive prompt.
Types
Always changelog relevant: feat, fix, perf
Others: docs, chore, style, refactor, test
Scope is optional.
The body should include the motivation for the change and contrast this with previous behavior.
The footer should contain any information about Breaking Changes and is also the place to reference GitHub issues that this commit Closes.
Don't care about developing the backend? Just want to test something against our production backend? We have yet another proxy for that:
# terminal 1
yarn yaproxy
# terminal 2
yarn dev:www
(Obvious )Warning: whatever you do here is for realz, if you login to your account and change things they are changed on republik.ch!
Install the ngrok cli: brew install --cask ngrok
Login to ngrok with ngrok authtoken <token> (You can find your token at https://dashboard.ngrok.com/auth)
After adding the authtoken, you must now add the following tunnels to your ngrok configuration file:
(Default config path is ~/.ngrok2/ngrok.yml)
tunnels:
republik-frontend:
proto: http
addr: 3010
hostname: republik.eu.ngrok.io
republik-backend:
proto: http
addr: 5010
hostname: api.republik.eu.ngrok.ioNow you must update the following environment variables:
API_URL=https://api.republik.eu.ngrok.io/graphql
API_WS_URL=wss://api.republik.eu.ngrok.io/graphql
FRONTEND_BASE_URL=https://republik.eu.ngrok.io # optional
COOKIE_DOMAIN=.republik.eu.ngrok.io
CORS_ALLOWLIST_URL=http://localhost:3003,http://localhost:3005,http://localhost:3006,http://localhost:3010,http://localhost:3000,https://republik.eu.ngrok.io
Start your frontend and api using:
yarn dev --filter=@orbiting/www-app... --filter=@orbiting/api-app...Now run yarn ngrok:start in a new terminal inside the workspace-root.
Your local development servers are now relayed to the following ngrok tunnels.
| local-address | ngrok-address |
|---|---|
| http://localhost:3010 | https://republik.eu.ngrok.io |
| http://localhost:5010 | https://api.republik.eu.ngrok.io |
With this you're now able to test payment-options (such as Apple Pay) that are only available in a secure context.
The environment variable SERVER is used to determine which app to build and run on deploy. If SERVER is missing the api app is run.
A heroku-prebuild script runs scripts/prune.sh which runs turbo prune with the correct scope and moved the pruned apps and packages to the root directory.
A heroku-postbuild script is used to add a Procfile for running the scheduler on heroku for the api app.