For the 10th Day of Advent of Code, we're asked to solve a matching braces problem. This is a common parser exercise, but it's made a bit more complex in this case by using multiple types of braces. Specifically, we're handling a series of lines that contain ( and ), but also < and >, [ and ], and { and }.
The story jazzes it up as being the code of our submarine's navigational computer, which consists entirely of braces in a sort of eldritch horror version of brainfuck, but that's mostly just a distraction.
We're told up front that most of the data lines are bad. None, in fact, are properly balanced. Some are corrupted by having an invalid character (say, a ] closing when a } closing was expected), while others are just incomplete. (Seriously, who wrote this submarine's computer system?)
The instructions say to ignore incomplete lines for part one and consider only corrupted lines. It's a safe bet that part two is going to have us consider the incomplete lines (spoiler alert: it does), so let's save time and consider both possibilities. The problem, then, is to find lines that are corrupted and find their first illegal character.
Then we want to turn the character into a point count and add those up, again mostly just to prove that we did it with a single final answer. Let's have a go at it.
Continue reading this post on PeakD.