Practical Coding Interview Question Bank?

September 27, 2017 10:16 PM

I'm looking for some interview questions with a practical / common library element. Most of the Leetcode/ HackerR<ank question sets have no practical knowledge component, and focus on abstract data structure algorithms. I do fairly well at these, but in interviews, the questions I get are more practical in nature. Is there a question bank in a book or website I can review?

I do alright at basic algorithmic questions, but as a seniorish devops / sysadmin, the questions I get asked to coderpad / whiteboard are seemingly more practical than CTCI questions lean. They often have a common library component, but are still timed. I want to get to a point where I'm Googling less for basics like how the python Requests lib works, and get better at this faster than one practical question per tech screen I get.

Example question / library pairings I've seen in the wild:

* implement pstree: os.walk()
* calculate the latency of a specific API endpoint: requests, datetime
* extract pertinent data from this log file: re

Companies seem to be pulling these sorts of questions from somewhere. Is there a question bank I can pull some examples from?

posted by pwnguin to Work & Money (1 answer total)

I got a question like that as part of a recent interview. In that case, they sent the questions home with me, so I had a chance to Google what I didn't know off the top of my head. (As it happened, I had previously implemented something very similar, so I didn't have to Google very much.) I suspect that they might be pulling random practical examples from their business in order to get more realistic questions and to avoid people who are memorizing algorithms and exercises. (They also might be biasing their hiring process in favour of specific experience and against potential, for better or worse.)

Since you appear to be focusing on Python, one exercise that might be useful is figuring out which 10 or 20 packages are the most popular on PyPI and either a) trying to implement them yourself based on their description and/or unit tests, or b) finding an interesting package that depends on one or more of the top 10 popular packages and trying to implement it. (How to figure out which packages are most popular on PyPI is an exercise in itself, one which I will leave to you. :-)

They told me that what got me the job was my code comments, good variable names, and unit tests, so that might be something you keep in mind if you get a take-home test. They said that nobody else did those things, which surprised me.
posted by clawsoon at 6:30 AM on September 28, 2017