Advent of Code 2017 in Kotlin

I talk about Python a lot here but at work we also write quite a bit of Java (with Spring of course). Since it hit 1.0 in 2016, I’ve been hearing more and more about Kotlin as a viable alternative to standard Java on the JVM.

At Pivotal we’re seeing clients start to use Kotlin because of its brevity compared to Java, the ease of use of existing Java based libraries, the adoption by Google for Android, and the (relatively new) ability to also target Javascript.

To find out what all the fuss is about, I decided to try out Kotlin for last month’s Advent of Code set of programming puzzles.

True to its reputation, it was easy to get a project started in Kotlin (using IntelliJ) and familiar tools like JUnit for testing are available without any trouble.

The puzzles in Advent of Code are of a particular type, mainly computational tasks with some algorithmic tricks often necessary for the more difficult second half of each challenge. So this was really a test of how well I could use Kotlin for these kinds of problems, rather than, say, doing a lot of data cleaning or creating an Android app. The well written problem descriptions lend themselves very well to a TDD approach as they almost always include examples that capture the complexity of the problem.

You can take a look at my solutions for the first 21 days of problems on Github. In the run up to Christmas, family preparations and celebrations took precedence over coding puzzles, but I’ll update the repo if I get time to complete the last few days.

It’s early days for my use of Kotlin, but so far I’ve found the language to be a nice mix of my favourite parts of Python (simplicity and flexibility) and Java (type safety and language specific tooling). Recently there has even been a movement to build a community of data scientists using Kotlin, and Thomas Nield’s videos are a good way to get started in this area.