An open source web application for building, managing, and playtesting Magic: the Gathering cubes.
If you are interested in contributing towards Cube Cobra, please read the Contribution guidelines for this project.
You will need to install NodeJS, Redis, and an IDE of your preference (I recommend VSCode). You can find the necessary resources here:
Node 20
NodeJS: https://nodejs.org/en/download/
Redis Server:
brew install redis
apt-get install redis
After installing redis, start the server. On mac, a shortcut to do this is brew services start redis
. You can seet the status with brew services list
.
Localstack provides a local emulation of AWS Services required to run CubeCobra including S3, DynamoDB and Cloudwatch.
You may follow the installation guidelines from the localstack site. The recommended setup involves running localstack in a docker container, which requires Docker Desktop as well.
brew install localstack/tap/localstack-cli
curl
command from localstackOnce localstack is installed, you can start the server in the background with the CLI: localstack start --detached
. You can see the status with localstack status
.
VSCode (strongly recommended, but not required): https://code.visualstudio.com/ ESLint Extension for VSCode: https://marketplace.visualstudio.com/items?itemName=dbaeumer.vscode-eslint Prettier Extension for VSCode: https://marketplace.visualstudio.com/items?itemName=esbenp.prettier-vscode
VSCode (with the ESLint and Prettier extension) is the recommended environment. When using this setup, make sure that your selected workspace is the root folder that you have cloned, this will ensure that the ESLint plugin can work with our linting rules. Prettier will automatically apply standard formatting to your code. Using these plugins will make adhering to the linting and code formatting rules significantly easier.
For the first setup, you will need to run:
yarn install && yarn build
yarn setup:local
This will:
If you are on Windows, you will need to set bash as your script shell:
You will need to make sure you have bash
installed somewhere and run the following command [with your bash
path in place of the path below].
yarn config set script-shell "C:\\Program Files\\git\\bin\\bash.exe"
Then you can start the program like so:
yarn start:dev
This script will:
You can now open up a browser and connect to the app through: http://localhost:8080.
(Despite the fact that node says it is running on port 5000 in the logs, you should use port 8080 to connect.)
Nodemon will restart the application anytime there is a change to a source file.
After accessing the application locally you will need to create a new user account using the "Resister" link on the nav bar.
Environment variables are populated from the .env
file. There is no .env
file checked in, so the setup script copies .env_EXAMPLE
to .env
and with some default values to support CubeCobra backed by LocalStack.
You can run a local instance of Cube Cobra against real AWS resources rather than LocalStack, if desired. After setting up S3, DynamoDB, and Cloudwatch using your AWS account, you can insert your credentials into the .env
file.
Here is a table on how to fill out the env vars:
Variable Name | Description | Required? | |
---|---|---|---|
AWS_ACCESS_KEY_ID | The AWS access key for your account. | Yes | |
AWS_ENDPOINT | The base endpoint to use for AWS. Used to point to localstack rather than hosted AWS. | ||
AWS_LOG_GROUP | The name of the AWS CloudWatch log group to use. | Yes | |
AWS_LOG_STREAM | The name of the AWS CloudWatch log stream to use. | ||
AWS_REGION | The AWS region to use (default: us-east-2). | Yes | |
AWS_SECRET_ACCESS_KEY | The AWS secret access key for your account. | Yes | |
CUBECOBRA_VERSION | The version of Cube Cobra. | ||
DATA_BUCKET | The name of the AWS S3 bucket to use. You will need to create this bucket in your account. | Yes | |
DOMAIN | The domain name of the server. Used for external redirects such as emails. | Yes | |
DOWNTIME_ACTIVE | Whether or not the site is in downtime mode. | ||
DYNAMO_PREFIX | The prefix to use for DynamoDB tables. You can leave this as the default value | Yes | |
EMAIL_CONFIG_PASSWORD | The password for the email account to use for sending emails. | ||
EMAIL_CONFIG_USERNAME | The username for the email account to use for sending emails. | ||
ENV | The environment to run Cube Cobra in. | Yes | \ |
NITROPAY_ENABLED | Whether or not to enable NitroPay, our ad provider. | ||
NODE_ENV | The environment to run Cube Cobra in. | Yes | |
PATREON_CLIENT_ID | The client ID for the Patreon OAuth app. | ||
PATREON_CLIENT_SECRET | The client secret for the Patreon OAuth app. | ||
PATREON_HOOK_SECRET | The secret for the Patreon webhook. | ||
PATREON_REDIRECT | The redirect URL for the Patreon OAuth app. | ||
PORT | The port to run Cube Cobra on. | Yes | |
REDIS_HOST | The URL of the Redis server. | Yes | |
REDIS_SETUP | Whether or not to setup the Redis server - this is needed for Redis but not for elasticache. | ||
SECRET | A secret phrase for encryption. You can leave the default value. | Yes | |
SESSION_SECRET | A secret phrase for session encryption. You can leave the default value. | Yes | |
SESSION | The name of the session cookie. You can leave the default value. | Yes | |
TCG_PLAYER_PRIVATE_KEY | The private key for the TCGPlayer API. | ||
TCG_PLAYER_PUBLIC_KEY | The public key for the TCGPlayer API. | ||
CACHE_ENABLED | Whether or not to enable caching. | ||
AUTOSCALING_GROUP | The name of the autoscaling group this instance is run in, used for the distributed cache. | ||
CACHE_SECRET | The secret for the distributed cache. |
In the initial setup scripts, yarn update-cards
is what creates the card definitions. Running this script will pull the latest data from scryfall.
If you want card analytics, can run the following script:
yarn update-all
This will, in sequence:
Express 4 provides a minimalist web framework to support both template rendering with PugJS 3 and definition of JSON-based API endpoints. HTML templates are mainly used to render a minimal page for React to bootstrap itself into with initial props injected from the server.
We keep all card definitions in large pre-processed files, so that nodes in production just need to download and load the files, and can fetch the latest files from S3 when they're ready. We do this because it's much faster to read from memory than to have to make requests to some other service anytime we need card data.
An external process is responsible for updating the card definitions, and uploading to S3. This same process is also responsible for updating the card analytics, and data exports.
We use redis for concurrency control for multiplayer drafting. All redis operations are handled in multiplayerDrafting.js
Each instance of the express server runs a job using node-schedule on a nightly basis to update the in-memory carddb from s3.
Bash scripts (jobs/definition
) are executed periodically on AWS to run hourly, daily & weekly jobs.
Card filters are defined that can be used by the frontend and backend. Nearley is a nodejs parser toolkit that is used to generate code that define filters that can be applied to the card database.
TypeScript 5.5 is gradually being rolled out to replace usage of vanilla JS components with PropTypes.
Components are provided by Reactstrap 9 which is powered by [Bootstrap v5.1][boostrap], which uses SCSS.
Components typically directly use bootstrap classes for additional styling, but may also use global classnames defined in public CSS files.