Best Available Seat

This repository my solution to the 'Best Available Seat' code challenge. A solution can be retrieved via API or the web. See below for information about local development and running tests.

This application has been deployed to Heroku: https://best-available-seat.herokuapp.com/

API Documentation

POST /api/seats - Find the best available seats in a venue.

Submit a POST request to the /api/seats endpoint to have the system determine the best available seats in the provided venue. The JSON payload must be structured like this:

{
    "requested_seat_count": 2,
    "venue": {
        "layout": {
            "rows": 10,
            "columns": 50
        }
    },
    "seats": {
        "a1": {
            "id": "a1",
            "row": "a",
            "column": 1,
            "status": "AVAILABLE"
        },
        "b5": {
            "id": "b5",
            "row": "b",
            "column": 5,
            "status": "AVAILABLE"
        },
        "h7": {
            "id": "h7",
            "row": "h",
            "column": 7,
            "status": "AVAILABLE"
        }
        // ...
    }
}

You can substitute your own requested_seat_count, venue or seats as necessary. Each seat must have a status of either 'BOOKED' or 'AVAILABLE'.

A successful response will look like this:

{
  "seats": [
    "b1",
    "b2",
    "b3"
  ],
  "error": ""
}

An error response will look like this:

{
  "seats": [],
  "error": "No row count specified."
}

The error value in the response will either be a string or an object representing a collection of errors.

GET /api/venue/{rows}/{columns} - Generate a JSON representation of a Venue

Submit a GEt request to the /api/venue/{rows}/{columns} endpoint to have the system generate a JSON description of a venue. The rows and columns must be integers. The system will not let you create a venue with more than 702 rows.

A successful response will look like this:

{
    "venue": {
        "layout": {
            "rows": 2,
            "columns": 2
        }
    },
    "seats": {
        "a1": {
            "id": "a1",
            "row": "a",
            "column": 1,
            "status": "AVAILABLE"
        },
        "a2": {
            "id": "a2",
            "row": "a",
            "column": 2,
            "status": "AVAILABLE"
        },
        "b1": {
            "id": "b1",
            "row": "b",
            "column": 1,
            "status": "AVAILABLE"
        },
        "b2": {
            "id": "b2",
            "row": "b",
            "column": 2,
            "status": "AVAILABLE"
        }
    }
}

Local Development

To run this application locally you must have Docker and Docker Compose installed on your host machine.

1. Clone this repository to your host machine.
2. Build the docker images: docker-compose build
3. Install the PHP dependencies: docker-compose run --rm php composer install
4. Install the Node dependencies: docker-compose run --rm node npm install
5. Build the assets: docker-compose run --rm node npm run local
6. Spin up the docker services: docker-compose up -d

You should now be able to reach the application by pointing your browser to http://localhost:8000.

To run the test suite, use docker-compose run --rm php vendor/bin/phpunit

Implementation Notes

• This application is built in PHP using the Laravel PHP framework. The web views are built on top of Laravel Livewire.
• The seat definitions in the JSON payload will represent all of the seats in the venue; any seat not listed in the seats object will not be available for analysis.
• The core functionality for finding the best available seats is defined in the app/Actions/FindBestAvailableSeats.php class.
• For the sake of this exercise I have assumed that groups of patrons will not be split across rows; if there are more patrons than there are seat columns available then the group will not be seated.
• I have also assumed that no venue will have more than 702 rows of seats. ('a' through 'zz'.)