Overview

Simple Discord bot for counting the number of votes in a given list of movies for the movie night on Discord.

Install

Dependencies can be installed into a virtual environment, with pip, using the commands

python3 -m venv <virtual_env_name>
pip install -r requirements.txt

or using the convenience provided in the makefile with

make install
source env/bin/activate

Usage

Requires an API key from Discord Developer Portal for usage inside of the os.getenv['API_KEY'] and a channel ID for the appropriate channel messages are to be placed into, this is retrieved via os.getenv['CHANNEL_ID'] as part of running the client.

Within the designated channel the commands are:

• !newvote movie 1, movie 2, movie 3, ... starts a new vote, only those with the role of 'Admin' can begin a vote.
• !vote X Y Z allows people to vote after a new vote has begun, the values correspond to the numerical value within list that is presented by the bot. Votes are weighted in order: the first vote is worth 3 points, the second is worth 2, and the last vote is worth 1.
• !changevote X Y Z someone can change their vote if it has already been cast, reusing !vote would stop them duplicating the vote.
• !endvote will end the vote casting and write the movie with the highest number of points into the channel.

Testing

Using GitLab CI for testing, this is activated via repository mirroring to GitLab and using a .gitlab-ci.yml file.

The stages for this pipeline are:

• test: run the bot_test.py file which runs another tester bot account that sends various messages into an isolated test channel, verifying the responses from the movie bot.
• build-image: this will build the latest image for the application, using the Dockerfile within .docker/bot.yml, and push it into AWS Elastic Container Registry (ECR). This occurs only on the master branch and uses the docker-in-docker service from GitLab.

Deployment

TODO: Update with some sort of blue/green etc.

At present, this is deployed using AWS Fargate, executing a single service for the container.

The deployment is achieved, after initially being setup through Terraform with terraform apply inside of the .tf directory, with a couple of ECS-related AWS CLI commands. These create a new task definition and update the running service that was created in Terraform, the latest task definition is then used on the running service to update to the latest image which was built in the build-image stage of the pipeline.