ticketfrei3/CONTRIBUTING.md

Kibicara Contribution HowTo

Setup Development Environment

General

1. Install python>=3.10 and development packages (apt install python3-dev g++ on Ubuntu)
2. Run ./setup.sh

Backend

0. cd backend
1. Activate your dev environment with source .venv/bin/activate
2. Install with pip install .
3. Turn off production mode: sudo su -c 'echo "production = 0" >> /etc/kibicara.conf'

Cheatsheet

• Install Kibicara with pip install .
• Execute Kibicara with kibicara (verbose: kibicara -vvv)
• Interact with Swagger REST-API Documentation: http://localhost:8000/api/docs
• Test and stylecheck with tox
• Fix style issues with black -S src tests

Frontend

1. Install node.js (e.g. via nvm)
2. cd kibicara-frontend
3. Install the dependencies with npm i
4. Install Angular with npm i -g @angular/cli
5. Turn off production mode if you have not already (see above in backend).
6. Start the backend in a different terminal
7. To serve and open the application, run ng s -o. The application will open under http://127.0.0.1:4200.

Creating an account

Without local mailserver setup

1. Start the server with kibicara -vvv.
2. Got to http://localhost:8000/api/docs. Use the POST /api/admin/register endpoint to create an account (minimal password size is 8).
3. Check the log output for a line that contains a link with the token:

<sometimestamp> kibicara.webapi.admin http://127.0.0.1:4200/confirm?token=<copythistoken>

4. Paste the token into the POST /api/admin/confirm/{register_token} endpoint at http://localhost:8000/api/docs.
5. Done! Try to log in with the Authorize button on the page or use the frontend

With local mailserver (e.g. openSMTPd)

1. Install the opensmtpd package and start it.
2. The mails should be usually delivered in ~/Maildir. Use your unix username as email address and register via frontend or manually at http://localhost:8000/api/docs.

Contribution Guidelines

Branches

• Development: The development branch is used to add new features.
  • Only rebase of feature branches is allowed.
  • On Release a release tag will be created
• Feature-Branches:
  • A feature branch will be used to develop a feature.
  • It belongs to one developer only and force push is allowed.
  • A rebase onto development is necessary to merge the feature. Code reviews are encouraged.

Write Tests

We use pytest as a test engine. It is executed by tox, so you need to run tox on the command line to run the tests.

Commit Messages

Commits should define small components. Please write your commits with the following pattern:

[core] Add censor for filtering messages #1312

You can use these tags:

• [core] Feature for Kibicara core
• [frontend] Feature for Kibicara frontend
• [$platform] Feature for platforms, e.g.
  • [twitter]
  • [telegram]
  • [email]
  • ...
• [tests] Tests
• [doc] Documentation
• [misc] e.g. github action files
• #\d+ if commit is related to specific issues or merge requests

Don't use tags which are not listed here. Don't start your commit message with lower case. Commit messages which do not fulfill these guidelines will not be merged into the development branch.

Comments

Python

We use pdoc3, which takes prose and comments automatically from the docstrings in the repository.

Use google style comments to secure correct display of your docstrings.

Please don't leave trailing whitespace or empty lines at the end of your files.

Merge Requests

The main development team does not need to make merge requests.

If you open a merge request, please include a description which explains the improvement.

Code reviews

Before a merge request gets rebased onto and merged into development, at least one person has to approve it; this also increases the number of people who know the code. So please request a review from someone from the core development team.

How to implement a new Platform/Social Network

tl;dr

1. Implement the following modules in platforms/<your-platform>/:

• bot.py
• model.py
• webapi.py

2. Import your bot in kibicara/webapi/__init__.py.
3. Generate the FastAPI stuff
4. Generate the angular components for the kibicara-frontend from the FastAPI stuff

Explanation

In kibicara/platforms/<your-platform>/bot.py, you write the functions through which the platform asks the social network for new messages, and publishes messages to the social network. You need to inherit the bot from the Censor class at kibicara/platformapi.py.

In kibicara/platforms/<your-platform>/model.py, you define a database layout. You will probably need to store the following things:

• authentication credentials,
• timestamps/IDs of the last seen message,
• recipients if the social network doesn't post publicly,
• platform-specific settings
• anything else your platform needs

In kibicara/platforms/<your-platform>/webapi.py, you can define HTTP routes. You will need them to:

• let admins authenticate to the social network in the kibicara web interface
• update platform-specific settings

To run the platform, you need to import the bot in kibicara/webapi/__init__.py.

Acceptance criteria for bots (Checklist)

A bot should have at least this functionality:

• Kibicara REST API (hood admins):

  • Endpoint for creating a bot associated with a hood
  • Endpoint for deleting a bot associated with a hood
  • Endpoint for updating a bot associated with a hood by id (The update endpoint may be optional if it does not provide any additional benefit to delete and create. An example where it does provide additional benefit is when there is a database which holds state like user ids in telegram. Then deleting the bot would also delete the user ids which would stop the bot from working.)
  • Endpoint for getting the bot configuration by id
  • Endpoint for starting and enabling a created bot by id
  • Endpoint for stopping and disabling a created bot by id
  • Endpoint for getting the status of a bot (INSTANTIATED, RUNNING, STOPPED)
  • Endpoint for getting all bots of a hood

• Bot functionality (user):

  • Possibility for a user to subscribe to the bot on their platform
    • e.g. Twitter itself already covers this by their follow/unfollow button
    • e.g. Telegram /start command
    • e.g. E-Mail REST Endpoint for entering and confirming an email
  • Possibility for a user to unsubscribe to the bot on their platform
    • e.g. Twitter itself already covers this by their follow/unfollow button
    • e.g. Telegram /stop command
    • e.g. E-Mail REST Endpoint for removing an email by token
  • Possibility for a user to send a message to the bot that gets broadcasted
    • e.g. Twitter via mentions or direct message
    • e.g. Telegram via direct message to the bot
    • e.g. E-Mail via e-mail to a specified address
  • Possibility for a user to get bot messages
    • e.g. Twitter via posts or retweets
    • e.g. Telegram via direct message from the bot
    • e.g. E-Mail via e-mail to the user's address

• Web Interface (hood admins)