kasznar / reboarding

Accenture competition first round

Reboarding to the office

Accenture competition second round

Running in docker

Build the image:
docker image build -t reboarding .

Run the docker image:
docker run -p 8080:8080 -d reboarding [ARGS]

To start the application basic configuration shall be provided. This includes the dates where the required distance between reservable seats decreases. These dates provided in the following format: 'YYYY-MM-DD'. Image file of the office layout are stored in the resources. Seat coordinates mapped in advance by hand and stored in seat-coordinates.json.

List of arguments:
--date.stepTo1
--date.stepTo2
--date.stepTo3
--date.stepTo5
--date.endOfPeriod --numberInWaitnigListBeforeNotification

Distribute seats

To pack the most people to the office with the given distance criteria, the problem can be represented as a graph. The nodes are connected when they are further apart from each other than the given distance. The graphs maximal clique is when the most seats can be reserved, and they all compile to the rules. This is find by an implementation of the Bron-Kerbosch algorithm. With this approach the application finds the optimal seating distribution in about 2 minutes.

Testing locally

For local testing dummy data is provided in the application.properties file, so it can be started without any arguments or configuration.

API testing

API testing is done by Postman Script. For these scripts to run correctly, the current date needs to have a corresponding capacity. For this reason the application can be started with apitest argument for local testing.

docker run -p 8080:8080 -d reboarding apitest
newman run postman/postman_collection.json -g postman/postman_globals.json   

CI/CD

Unit tests and build run by Github Action, after every commit or pull request to master.

Endpoints

HTTP Method	URL	Description
POST	/api/reboarding/register?userId={userId}&day={day}	Register a user to a given date
POST	/api/reboarding/entry/{userId}	Check in a user for today
POST	/api/reboarding/exit/{userId}	Check out a user for today
GET	/api/reboarding/status?userId={userId}&day={day}	Get the status of the user for a given day
GET	/api/reboarding/layout?day={day}	Get the status of all seats for a given day (image)
GET	/api/reboarding/layout?userId={userId}&day={day}	Get the location of the reserved seat on a given day (image)