mnigh / ghog-day

groundhog day website 🦫

GROUNDHOG-DAY.com

It's a real holiday.

About

GROUNDHOG-DAY.com is the leading data source for North America’s prognosticating groundhogs and their yearly predictions. It's a fun little express app with a frontend you can browse and an API you can use. Historical Groundhog Day data is collected from a variety of sources, including an academic paper.

• The frontend is pretty straightforward: nunjucks and sass and JS only when I have to.
• The API does what APIs do: serves resources in JSON. It's free, fast, and simple.

There's kind of an interesting concept here: the data is basically static and it's all about GETs. To make it cheap, fast, and easy, there's one container with an in-memory DB that is seeded when the container boots up on deploy. Super easy to work on and cheap to run, even with 100s of thousands of hits a month.

Using the API

Please get in touch if you are using the API and you need something, because I can probably make it work better if I have enough real-life use-cases.

Read more on the API.

There's an OpenAPI spec at Groundhog-Day-API.v1.yaml and a SwaggerHub page where you can test the endpoints.

Getting started

Install npm

npm is a javascript package manager. It downloads project dependencies and runs node applications.

You'll need node version v14 or higher to run the app. (If you have an M1 Mac, you'll need v15 or higher.)

Install docker

A docker container allows a developer to package up an application and all of its parts. This means we can build an app in any language, in any stack, and then run it anywhere — whether locally or on a server.

Build and run with npm

Just install the dependencies and run it. Pretty slick. 😎

# install dependencies
npm install

# run application in 'dev' mode
# (ie, the server restarts when you save a file)
npm run dev

# run application in 'prod' mode
npm start

The app should be running at http://localhost:3000/.

On a Mac, press Control + C to quit the running application.

Run tests with npm

# run unit tests
npm test

# run linting
npm run lint

unit tests

The unit tests are pretty straightforward. I'm using supertest to make requests and then cheerio to load + parse the responses and then we can make assertions against them. Using cheerio, we can load in a string like "<main><p>hello</p></main>" and then traverse it using jQuery selector syntax. So we can write assertions against stuff like $('main > p').text(), which is far better than string equality-type stuff.

There's not so many tests at the minute, but I am getting an A+ on LGTM which — dear reader — is the highest possible mark.

Build and run as a Docker container

# build an image locally
docker build -t pcraig3/ghog:<tag> --build-arg GITHUB_SHA_ARG=<tag> .

# run the container
docker run -it -p 3000:3000 pcraig3/ghog:<tag>

The container should be running at http://localhost:3000/.

On a Mac, press Control + C to quit the running docker container.

Push to Cloud Run

The main.yaml file contains instructions to deploy the service to Cloud Run, or you can use the CLI command below to do it manually.

gcloud run deploy {SERVICE} --project {PROJECT} --image gcr.io/can-hols/hols:{TAG}