Skip to content

G-Leopard/flights-assignments

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Flights assignment

Solution to the take-home assignment of an interview process I took part in 2022. Description of the assignment can be found here.

System spec

All the work has been done on a 2021 16" Apple Macbook Pro with the following specs:

  • M1 Pro processor (10-core CPU and 16-core GPU)
  • 32 GB of RAM
  • 1 TB SSD

Software

  • Go 1.18.3 (with chi 5.0.7)
  • macOS Monterey 12.4

Considerations

Algorithms and benchmarks

Before developing the server, some algorithms were developed and benchmarked. A compromise between speed, efficiency, and readability led to the choice of the algorithm used in the server.

Algorithms and benchmarks can be found under cmd/cli.

Benchmark results: Benchmark results

The algorithm chosen is sliceDiff.

Sorting and non-duplicates

As far as I understood, sorting the list of flights was a suggestion, not a requirement. All the algorithms implemented don't sort the flights: they look for non-duplicated elements.

Track() contains the algorithm used by the server.

API Format

The API has two endpoints:

/health

Basic health check to detect if the application is responding. It answers to GET requests only.

Request example

curl -i http://localhost:8080

Response example

HTTP/1.1 200 OK
Date: Sat, 09 Jul 2022 17:51:45 GMT
Content-Length: 0

/track

It returns the starting airport and the final airport from a list of unordered flights.

It answers to POST requests only. The list of flights should be fed via .json.

Json example

[
    {
        "start": "SFO",
        "end": "ATL"
    },
    {
        "start": "ATL",
        "end": "GSO"
    }
]

More examples can be found under testdata.

Request example

curl -i \
    -X \
    POST \
    -d \
    @testdata/long.json \
    http://localhost:8080/track

Response example

HTTP/1.1 200 OK
Content-Type: application/json
Date: Sat, 09 Jul 2022 18:01:21 GMT
Content-Length: 36

{
  "start": "BGY",
  "end": "AKL"
}

How to run the server

Prerequisite: you need either Golang or Docker installed on your machine.

The server runs on port 8080.

Go

Before running the server you might want to execute go mod tidy to download potentially missing dependencies.

  • make http runs the server on your machine

Docker

The docker-compose can be found here. The docker-compose builds the Docker image corresponding to this Dockerfile and starts the server.

  • make dev runs the docker-compose in the background
  • make logs follows the docker-compose logs
  • make nodev stops and removes the container
  • make devrebuild rebuilds the image and runs the container in the background

Execute requests

The Makefile also contains commands to execute HTTP requests to the server's endpoints.

  • make health performs a GET request to the /health endpoint
  • make track performs a POST request to the /track endpoint. It uses the file long.json. You can change the input file by feeding the variable DATA to the command. Example:
DATA=testdata/short.json make track

Run benchmarks

You can run the benchmarks on your machine (you need Golang installed) with the command: make bench.

Tools

  • trivy was used to identify vulnerabilities in the Dockerfile
  • hadolint was used to lint the Dockerfile
  • golangci-lint was used to lint the Go code

Potential improvements

The following are ideas that haven't been implemented due to lack of time.

  • input data validation: output can be unpredictable if the list of flights is not valid
  • config file: it comes handy having the HTTP port configurable. go-yaml is a good library for .yaml config files
  • logger: logging has been implemented with the standard library's log package. A better alternative would be logrus
  • data generator: the idea was to create a function able to generate a list with hundres/thousands of flights, to better understand which algorithm scales better
  • live reload: reflex is a great tool to test code changes without having to restart the server
  • unit tests: self-explanatory