cla-bot

A GitHub bot and Web UI for managing contributor license agreements.

Installation

This instance is forked from the original and ran by the Python Software Foundation for python/cpython.

It's deployed on Heroku via https://git.heroku.com/cpython-clabot.git with configuration stored in heroku.yml.

The default Dockerfile here includes the frontend CLA bot Web app and is the same as the upstream CLA bot. Since it's AWS-specific, we use a separate file: Dockerfile.cpython. This one includes both EdgeDB Server and the CLA bot Web app. This is because Heroku doesn't support network linking of Docker images, and the only network protocol exposed for EdgeDB would be HTTPS which is insufficient for our needs.

Since EdgeDB here is backed by a Postgres instance, there can be multiple frontend EdgeDB server processes without an issue.

Note that Dockerfile.cpython only includes files from the upstream CLA bot Web frontend image hosted on hub.docker.com instead of building the files inline. This is because the upstream build uses COPY . . which invalidates on any non-ignored file change in this repository. And the Yarn Next.js build takes over 5 minutes.

Maintenance

When upgrading CLA bot frontend code, rebuild ambv/cla-bot-frontend by running:

docker build -t ambv/cla-bot-frontend .
docker push ambv/cla-bot-frontend

This is a helper image that is merged with the EdgeDB server later using Dockerfile.cpython. That main image is built by Heroku itself when you push to the Heroku git from this repo. You can also rebuild ambc/cla-bot-cpython manually by running:

docker build -t ambv/cla-bot-cpython . -f Dockerfile.cpython

You can run a local container and use it in a regular fashion granted that there's an .env file with DATABASE_URL in it from Heroku Postgres. You can read it by running heroku config. Once the file is there, start the container with:

docker run -it --rm --name cla-bot --env-file .env -p 3000:3000 -p 5656:5656 ambv/cla-bot-cpython

Running a deploy on Heroku

$ git clone $THIS_REPO
$ heroku git:remote -a cpython-clabot
$ git push heroku cpython-clabot:main

This will output build information during the push.

What follows is the original README.

Development

Requirements:

Node.js v13.3.0
yarn (install with npm install -g yarn)

A full setup requires also:

a configured OAuth application in GitHub
a configured GitHub application in GitHub
an EdgeDB instance configured with schema from dbschema/
.env file populated with proper application settings
a web hook for pull requests

For more information on this first-time configuration, refer to the documentation in the project Wiki, at Configuration.

Getting started:

# install dependencies
yarn install

# run the development server
yarn next

Project structure

This project uses onion architecture, with the following namespaces:

constants contains configuration constants
components contains reusable React components
pages is a folder handled by Next.js, with routes: pages and api
pages-common is a folder containing common code for pages and api used in Next.js front-end
public is a folder handled by Next.js, containing static files
service.domain contains domain objects and interfaces for external services
service.data contains concrete implementations of external services
service.handlers contains business logic

Business logic is lousy coupled with the data access layer, since it is only aware of interfaces, not concrete implementations of DAL logic. Everything inside the service folder is abstracted from Next.js and should be reusable with other web frameworks, unmodified.

Documentation

For documentation, refer to the project Wiki.