Coding Challenge

Background

Plantix Community is a social network of plant experts, supporting each other in diagnosing plant diseases. Each expert has a unique id, a list of plant species (topics) they can give advice on and a list of other experts they follow:

------------------------------------
  PlantExpert
------------------------------------
    uid:          String
    plants:       List[String]
    following:    List[String]
------------------------------------
  fetch(uid: String): PlantExpert
------------------------------------

The fetch(uid) API is already implemented and available in your Python SDK (plantix.py). Feel free to re-implement this code if you need to use a different programming language.

The challenge

Suppose an expert with uid="0" exists. Can you find what topics (plants) are covered by the experts in his/her entire network of connections (any degree)?

1. Implement a new SDK method in plantix.PlantixApiClient, which returns a sorted Tuple[String] of the n most covered plants in the network of experts reachable from expert start. For example, if there are 5 experts on tomato, 3 on banana, and 10 on cucumber - all reachable from user with uid="3", this method should return:

   >>> from plantix import *
   >>> PlantixApiClient().find_topics(start="3", n=2)
   ('cucumber', 'tomato')
   

   in this order.

2. Write unit- and integration tests for your solution.

3. Run your solution with input start=0, n=3. That is, calculate the top 3 most covered plant species in the network of experts reachable from uid=0.

4. Analyze the running time and memory complexity of your solution. Consider best- and worst-cases.

5. Write a small design doc to explain the problem and your approach to solving it. Add your O() analysis in there too.

6. Type in the output from (3) as an answer in your application form and attach a single zip archive containing all of your source code (1-2) and design doc (3-5) for review. For example, if your program produced the output ('cucumber', 'tomato'), you need to type in cucumber, tomato in the application form and attach a zip file with the SDK, your code, tests, and design doc.

Requirements

• This is a small task. If you find yourself writing a lot of code, you might want to consider a different approach.
• Use any modern language you like; however we only provide an "SDK" for Python. You can use the standard library of your language, but no other packages. Your code should compile without any dependencies.
• Write production-quality, correct, clean, maintainable, testable, tested, documented and scalable code.
• Feel free to remodel PlantixApiClient, add new functions, methods, properties or classes if necessary. You can't modify PlantExpert however.
• You do not have unlimited amounts of RAM. Your solution should scale.
• Format your code the way you normally would when working in a team. If using Python, follow PEP8 and provide type annotations and docstrings.