bookbub / takehome Goto Github PK

Thank you for putting your time into taking this challenge and exploring the opportunity with BookBub. Make your best effort to package and structure your program to allow for understandability and future maintenance—that is, write this code with the same quality as a feature you'd be shipping to bookbub.com. There's no major gotchas or brain-teasers involved here, we're primarily interested in getting a sense for your practical ability to write code and solve an almost real-world problem. We don’t expect you to spend more than a small handful of hours on this, so feel free to timebox yourself to two or three hours. If you run out of time, you can document in the readme what you would have done with more time.

We want to build a tool to help BookBub’s editorial team quickly go through a list of books and identify the correct genre for each book. The approach we’re taking is to analyze the description of each book and calculate a score based on the presence of certain genre-specific keywords.

For each genre, we have several keywords that we expect to be present in the book description for that genre. We have assigned a point value to each keyword because some keywords provide stronger signals than others. The genre-fit score is calculated as:

total num keyword matches * avg point value of the unique matching keywords

Example Book:

{
  "title": "Hunger Games",
  "description": "In a not-too-distant future, the United States of America
    has collapsed, weakened by drought, fire, famine, and war, to be
    replaced by Panem, a country divided into the Capitol and 12 districts.
    Each year, two young representatives from each district are selected
    by lottery to participate in The Hunger Games. Part entertainment,
    part brutal intimidation of the subjugated districts, the televised
    games are broadcasted throughout Panem as the 24 participants are
    forced to eliminate their competitors, literally, with all citizens
    required to watch. When 16-year-old Katniss's young sister, Prim, is
    selected as the mining district's female representative, Katniss
    volunteers to take her place. She and her male counterpart, Peeta,
    the son of the town baker who seems to have all the fighting skills
    of a lump of bread dough, will be pitted against bigger, stronger
    representatives who have trained for this their whole lives. Collins's
    characters are completely realistic and sympathetic as they fight and
    form alliances and friendships in the face of overwhelming odds; the
    plot is tense, dramatic, and engrossing. This book will definitely
    resonate with the generation raised on reality shows like 'Survivor'
    and 'American Gladiator.' Book one of a planned trilogy."
}

Example Keywords CSV:

Genre, Keyword, Points
action, fast paced, 7
action, distant future, 4
action, fight, 6
mystery, murder, 5
mystery, death, 8
mystery, explosive, 4
sci-fi, future, 8
sci-fi, spaceship, 9

In the example above, the action score for Hunger Games is 15. Keywords distant future and fight both match and their average points are 5. With 3 total keyword matches, the action score is 15, with 3 * 5 = 15. The sci-fi score is just 8, since future is the only sci-fi match. The mystery score is 0 since there are no keyword matches.

Write a command-line program that takes in two arguments:

• A json file containing a list of books (sample json file)
• A CSV with genre/keyword/value rows (sample csv file).

Your program should print out the book titles alphabetically, and underneath each book title should be the three highest scoring genres with a score greater than 0 (including the score for that genre).

When run against the provided book and genre input files, your program sholuld produce the following output:

Hunger Games
action, 15.0
sci-fi, 8.0

Infinite Jest
literary fiction, 12.0

There are a few formatting rules which will help us grade your submission properly.

1. Include a space between the comma and the score when printing genres-point combinations

   Good

   action, 15.0
   

   Bad

   action,15.0
   sci-fi,       8.0
   

2. Include a new line after each book + score section and no where else.

   Good

   Hunger Games
   action, 15.0
   sci-fi, 8.0
   
   Infinite Jest
   literary fiction, 12.0
   
   

   Bad

   Hunger Games
   
   action, 15.0
   
   sci-fi, 8.0
   

That's it!

When you have finished building your program, please use our analysis tool that's included in this repo. This tool helps us grade your submission and also helps ensure your program is functioning as expected.

To use our tool, you will need to have a working node environment. Read here for how to install node with your favorite package manager, or here for building from source.

Once node is installed, run the following from the command line.

$ node bookbub_analyzer.js <code trigger>

The <code trigger> is the series of strings which is used to invoke your program, minus the command line arguments.

For example, if you invoke your program as follows.

$ python genre_detector.py books.json genres.csv

then please run

$ node bookbub_analyzer.js python genre_detector.py

If your program does not produce the correct output for the sample test case, the analysis tool will print an error message. Continue working on your program until the tool determines your output is correct. Once your program works as expected, the tool will generated a zip-file named bookbub_analysis.zip. Please include this file in your final submission.

If you experience any difficulties installing node or using our tool, don't sweat it. Just let us know when you submit and we'll be sure to inform the engineer grading your submission.

Please create a zip file containing

1. Any source code files needed to run your program

2. The bookbub_analysis.zip file that was generated

3. A README.md explaining:

   • Any steps necessary to run your program
   • Any interesting trade-offs or edge cases you ran into (this can really strengthen a submission)
   • Approximately how long you spent (this is not timed, but it’s helpful for us)

You can write your program in any language you like as long as we can run it on Linux. Here are a few qualities we expect to see in a good solution.

• Readbility
• Logical separation of concerns
• Effective use of data-structures

The purpose of this exercise is for you to show us your ability to write clean, non-trivial code.