<Advent of Code 2021 day 10> :thread:
# advent-of-codea
Advent of Code 2021 day 10
12/10/2021, 5:00 AMd
David Whittaker
12/10/2021, 5:25 AMI'm feeling like we've done something like this before in prior years.
s
Simon Nyström
12/10/2021, 5:52 AMday 1 2015 deals with counting parentheses at least
Simon Nyström
12/10/2021, 5:53 AMm
Marcin Wisniowski
12/10/2021, 6:05 AM💯 1
👍 1
n
nkiesel
12/10/2021, 6:15 AMMy approach was to repeatedly remove direct pairs from the line until it no longer changes. After that, removing all opening chars gives the char for part 1, and processing the remaining chars in reverse order gives the solution for part 2. Not the most efficient way, but still runs in < 100ms. Details in https://github.com/nkiesel/AdventOfCode2021/blob/Norbert/src/test/kotlin/Day10.kt
m
Michael Böiers
12/10/2021, 6:22 AMNot very pretty. Got the job done, but took me very long since I took a completely wrong and stupid approach first. Should prepare coffee for the next challenge. 🙂 https://github.com/MikeEnRegalia/AdventOfCode2021/blob/main/kotlin/src/main/kotlin/Day10.kt
d
David Whittaker
12/10/2021, 6:51 AM@Marcin Wisniowski - ooh TIL
mapNotNullK 1
David Whittaker
12/10/2021, 6:52 AM@Michael Böiers - I also used
whend
Dan Fingal-Surma
12/10/2021, 7:12 AMDan Fingal-Surma
12/10/2021, 7:14 AMSealed interface worked nicely here IMO
Dan Fingal-Surma
12/10/2021, 7:17 AMGot a wrong answer on part 2 first due to integer overflow
🙌 3
t
Tobias Suchalla
12/10/2021, 7:24 AMI used sealed classes as well. And ran into the overflow as well...
Anyway, I'm actually really happy with my solution.
https://github.com/Sydenth/aoc-21/blob/main/src/year2021/day10/Day10.kt
e
ephemient
12/10/2021, 7:47 AMhttps://github.com/ephemient/aoc2021/blob/main/kt/src/commonMain/kotlin/com/github/ephemient/aoc2021/Day10.kt no fancy data structures or anything, just a manual stack
p
Paul Woitaschek
12/10/2021, 8:50 AMm
Michael Böiers
12/10/2021, 8:50 AM@ephemient
private val pairs = "([{<".zip(")]}>").toMap()👀 3
k
kqr
12/10/2021, 9:45 AMm
Michael Böiers
12/10/2021, 11:02 AMHere’s my improved solution, had to try to implement it with a stack after having realized my error. 🙂 https://github.com/MikeEnRegalia/AdventOfCode2021/blob/main/kotlin/src/main/kotlin/Day10Proper.kt
👍 1
g
gnu
12/10/2021, 2:45 PMm
Michael de Kaste
12/10/2021, 2:48 PMwish I could have used the sealedmap as I said elsewhere in the channel, but this works fine too.
Single parse over the input to split the lines into incomplete and illegal versions and then do the proper math
https://github.com/mdekaste/AdventOfCode2021/blob/main/src/main/kotlin/year2021/day10/Day10.kt
p
Paul Woitaschek
12/10/2021, 2:50 PM@Michael de Kaste what's a sealed map?
👀 1
m
Michael de Kaste
12/10/2021, 3:27 PMwell it doesn't exist yet, thats the problem 😄
p
Paul Woitaschek
12/10/2021, 3:27 PMYeah right but what should it do?
m
Michael de Kaste
12/10/2021, 3:44 PMA sealed class knows all its subclasses because of the promise that you can only add those subclasses within the same file. This means that we could technically treat a sealed class as an enum. If you have a big list of sealed class/interface, you know that the elements of the list are ALWAYS a subtype of that sealed class/interface. You could always do
list.filterIsInstance(YourSealedClass.SubType1::class.java)l
Luke
12/10/2021, 4:26 PMLittle bit of Arrow's
Eitherp
phldavies
12/10/2021, 9:58 PMhttps://github.com/tKe/aoc-21/blob/main/kotlin/src/main/kotlin/year2021/Day10.kt another day, another enjoyable task.
😍 1
k
Kiet
12/11/2021, 10:28 PM3 Views