Fare Estimator

A fare estimator app which receives cab positions/paths for each ride in a csv file and returns a csv file with estimated fares for each ride. It uses Haversine distance formula to calculate distance between two (lat, lng) pairs.

Constraints

Input paths.csv consist of:
    id_ride, lat, lng, timestamp

Output result.csv will contain:
    id_ride, fare_estimate

Discard invalid entries:
    Filtering should be performed as follows: consecutive tuples p1, p2 should be used to calculate the segment’s speed U. If U > 100km/h, p2 should be removed from the set.

Fare estimation:

Amount when cab is moving at > 10km/h during day time (05:00, 00:00)
    0.74 per km

Amount when cab is moving at > 10km/h during night time (00:00, 05:00)
    1.30 per km

Amount when cab is idle <= 10km/h at anytime
    11.90 per hour of idle time

Standard flag amount is charged to the rides fare
    1.30, the minimum ride fare should be at least 3.47

Architecture

I could have gone with Golang's standard project layout, however, I chose Domain Drive Design (DDD) whereby each layer has its own responsibility, considering scalability factor, it seemed a great fit.

Directory structure

├── bin
│   └── fareestimatorcli
├── cmd
│   └── cli
│       ├── main.go
│       └── main_test.go
├── app
│   ├── domain
│   │   ├── fare
│   │   │   └── fare.go
│   │   ├── path
│   │   │   ├── path.go
│   │   │   └── path_test.go
│   │   └── ride
│   │       └── ride.go
│   ├── infrastructure
│   │   └── filesystem
│   │       └── csv
│   │           ├── mocks
│   │           │   ├── Reader.go
│   │           │   └── Writer.go
│   │           ├── reader.go
│   │           ├── reader_internal_test.go
│   │           ├── reader_test.go
│   │           ├── writer.go
│   │           ├── writer_internal_test.go
│   │           └── writer_test.go
│   └── usecases
│       ├── fare
│       │   ├── mocks
│       │   │   └── Service.go
│       │   ├── service.go
│       │   ├── service_internal_test.go
│       │   └── service_test.go
│       ├── path
│       │   ├── mocks
│       │   │   └── Service.go
│       │   ├── service.go
│       │   └── service_test.go
│       └── ride
│           ├── mocks
│           │   └── Service.go
│           ├── service.go
│           ├── service_internal_test.go
│           └── service_test.go
├── testdata
│   ├── coverage.out
│   ├── paths.csv
│   ├── paths_for_test.csv
│   ├── result.csv
│   └── result_for_test.csv
├── go.mod
├── go.sum
├── Makefile
└── README.md

What's includes in each layer?

• Domain: include models entities etc
• Usecases: include services
• Infrastructure: include infra related stuff, e-g: fs, mappers, repositories, clients etc

Setup

Clone this repository somewhere on your computer and make sure you have at-least go1.16 installed.

git clone [email protected]:muzfr7/fare_estimator.git

To download dependencies, change directory into cloned fare_estimator repository and run following command

go mod download

make sure you have paths.csv file in testdata directory, because it will be ingested directly for fare estimation.

Usage

Before proceeding further, make sure to change directory into cloned fare_estimator

Build and run

make run

this will build and run ./bin/fareestimatorcli ingesting ./testdata/paths.csv file, to produce ./testdata/result.csv file containing estimated fares for each ride.

Run tests

make tests

this will also generate ./testdata/coverage.out file, issue a make coverage command to view it in browser.

Cleanup

Issue following command to remove all generated files: ./bin/fareestimatorcli, ./testdata/result.csv and ./testdata/coverage.out

make clean

Notes

• Layered architecture is used so that this app can scale if more resources and time is added.
• Handler in cmd/cli/main.go file could be moved to presentation layer instead.
• I think a more scalable version of this could be developed using messaging queue instead of CSV file.
• Fare estimation service has O(n²) complexity which can possibly be reduced, however, I could not given time constraint!
• This solution is missing proper logger integration and tracker etc.
• Test coverage could be improved.