mayurshivakumar / velocitylimits Goto Github PK

Each attempt to load funds will come as a single-line JSON payload, structured as follows:

{
  "id": "1234",
  "customer_id": "1234",
  "load_amount": "$123.45",
  "time": "2018-01-01T00:00:00Z"
}

Each customer is subject to three limits:

• A maximum of $5,000 can be loaded per day
• A maximum of $20,000 can be loaded per week
• A maximum of 3 loads can be performed per day, regardless of amount

As such, a user attempting to load $3,000 twice in one day would be declined on the second attempt, as would a user attempting to load $400 four times in a day.

For each load attempt, you should return a JSON response indicating whether the fund load was accepted based on the user's activity, with the structure:

{ "id": "1234", "customer_id": "1234", "accepted": true }

You can assume that the input arrives in ascending chronological order and that if a load ID is observed more than once for a particular user, all but the first instance can be ignored. Each day is considered to end at midnight UTC, and weeks start on Monday (i.e. one second after 23:59:59 on Sunday).

Your program should process lines from input.txt and return output in the format specified above, either to standard output or a file. Expected output given our input data can be found in output.txt.

You're welcome to write your program in a general-purpose language of your choosing, but as we use Go on the back-end and TypeScript on the front-end, we do have a preference towards solutions written in Go (back-end) and TypeScript (front-end).

We value well-structured, self-documenting code with sensible test coverage. Descriptive function and variable names are appreciated, as is isolating your business logic from the rest of your code.

• Replace in memory cache by a persistent cache.
• Dependency injection sample service. Would be nice to mock out other dependencies.
• TODO comments should be worked out to further improve the code.

• Adding worker pool.
• Right now there are three go routines running for each stage(read request, process request and write response). Multiple workers/go routines can be added at appropriate step(s).
• Refactor the go routines to steps and write tests for them.
• Improve error handling.
• Improve file handling.
• Fix the file paths for input, config and output.
• Run profiler(pprof).