X-Lab Engineering interview challenge

The quickest and easiest way to get started is to clone the repo on a machine where Docker is installed and start it up by running ./scripts/start.

The application can be run without Docker, further information is below.

The application is containerised for Docker and can be run via docker or docker-compose commands. Docker will need to be installed to use this option.

The application can also be run without Docker, for this Node.js needs to be installed.

The easiest option is to use the convenience script (scripts/start) which includes the commands to start the application via docker-compose.

The commands to run the application with just docker are:

docker build . --build-arg NODE_ENV=development -t beer-quest

docker run --init -p 3000:3000 -p 9229:9229 -t beer-quest

Once the repo has been cloned, the dependencies need to be installed by running npm install in the project directory. The application can then be started by running npm run start or npm run start:watch.

By default, ports 3000 and 9229 are used by the application. If these are already in use the application will fail to start. Please ensure both ports are not in use before trying again.

Tests for the application can be run in one of two modes via the convenience script (scripts/test). In order to run the tests once, simply run the script with no options i.e. ./scripts/test. The tests can also be run in watch mode by passing -w to the script ./scripts/test -w. This will continually run the tests when file changes are detected and is a good option to use whilst developing the application.

A version of the application has been (manually) deployed to Azure Container Instances (ACI) and is available at http://beer-quest.uksouth.azurecontainer.io:3000/venues.

The deployment can be rerun using the script deploy.

The deployment can be deleted using the script delete.

ACI isn't the most feature rich product. HTTPS appears to require some additional configuration, hence HTTP! Port mapping also doesn't appear to be a working feature, hence the specific port.

Images are pushed to hub.docker.com/repository/docker/st3v3nhunt/beer-quest on Docker Hub for every commit to main.

• Some venues appear to be closed, these should probably be filtered out from the dataset either preload or during the load.
• The dataset is so small it doesn't seem worth introducing a database, obviously this would be different is the size was larger or the data was dynamic.
• The code for filtering venues based on scores has been kept simple. It would be possible to not run a filter if the rating had not been set however, as each filter operation should have a time complexity of O(n) I decided the additional code complexity wasn't worth the minimal performance hit. Given a larger datasets this might change but then the filtering would likely be done via a database.
• Some consideration has been given to malicious input, specifically the ratings and include distance params have some level of validation, however, the coords are not in anyway checked. This needs to be worked on.
• Ratings are taken as a minimum as I figured you'd want to have the good venues returned.
• The API could do with being documented so clients would have an easy time using it.
• It would be good to have tags create images with the appropriate tag.
• Each PR should create a review environment for testing prior to merge to main.