Advent of Code 2023 Day 1: Trebuchet?!

An unfortunate start, but some good lessons from today for the future.

(Disclaimer: I receive commissions on paid memberships through my link, but Codecrafters is not a sponsor of this channel, does not endorse any of my content, and does not review any of my videos.)

[hyper-neutrino]

For part 2, it's actually probably simpler to check if the line starts with one of our match conditions (a word or a digit) and then shift by one character each time. The regex is how I did it, but in hindsight, I think it's more complicated than necessary and I think manually scanning the line would be much smarter.

This is basically just how the regex does it under the hood but it's probably easier to understand and doesn't require you to understand (?=...).

Also, I got the name of that wrong, sorry. It's a lookahead assertion ("this thing is ahead, but don't match it") not a non-capturing group (completely unrelated).

[nonyabizness]

The edge cases on `oneight`, `eightwo`, etc. were something i missed considering so got stuck for a while on this problem thanks a lot! :D

[xdf]

Great solution. Much cleaner than mine. I replaced "one" with "one1one" for each number, and then removed non-digits. This preserves the text for overlaps, but still gets the digits in the right order. I think your solution is cleaner.

For future videos, can you make your font way bigger. It was nearly impossible to read on a phone.

[nox]

D = {'one' : 1, 'two' : 2, 'three':3, 'four':4, 'five':5, 'six':6, 'seven':7, 'eight':8, 'nine':9, '1':1, '2':2, '3':3, '4':4, '5':5, '6':6, '7':7, '8':8, '9':9, '0':0}
that's what i initially/intuitively did /// geez 1st day

[ZyroZoro]

Thank you for your videos! I learn a lot from them.

[felix.delrubio]

On the "eightwothree" entry, I was of the opinion that the two didn't exist because the "t" was consumed by the eight. It gave me the right answer.

[_bergee_]

What I did in second part was:

-> generated all overlapping digit words combinations (like twone, oneight) and the corrected ones
-> replaced the ovelapping words with corrected ones
-> replaced all the words with numbers

[eavdmeer]

I bypassed the attempt at frustrating things with the overlapping words by searching for the first string index and last string index of all pairs '1' and 'one', '2' and 'two' and so on. I keep track of the highest and lowest index and value during the loop.
Not sure if Python has a 'lastIndexOf()' kind of thing

[nardove]

Thank you for sharing! it will be great if you could increase the font size for the upcoming videos, is very difficult to see clearly specially with that typeface

[batlin]

I was stuck on part 2 for exactly the reason you described for around 10 minutes before someone on chat gave me a hint to try the testcase "eightwo", which my code of course transformed to "8wo". That was pretty annoying indeed that the testcase didn't exercise that behaviour since stepping through the full input to look for possible issues was a bit impractical and it wasn't obvious behaviour.

[RickMMarina]

Hi, thanks for the video I didn't know the possitive lookahead pattern.

Anyways, I got the correct solution just normalazing the string for special cases:
private string Normalize(string cad) {
Dictionary<string, string> replaces = new() {
{"oneight", "oneeight"},
{"eightwo", "eighttwo"},
{"eighthree", "eightthree"},
{"threeight", "threeeight"},
{"twone", "twoone"},
{"sevenine", "sevennine"},
{"nineight", "nineeight"},
{"fiveight", "fiveeight"},
{"one", "1"},
{"two", "2"},
{"three", "3"},
{"four", "4"},
{"five", "5"},
{"six", "6"},
{"seven", "7"},
{"eight", "8"},
{"nine", "9"}
};
foreach (var kv in replaces) {
cad = cad.Replace(kv.Key, kv.Value);
}
return cad;
}

[jtfidje]

Maybe they chose to limit the test data in that way to "slow down" anyone using LLMs?

[xavierjulien]

Please zoom in, thx for your content

[murphygreen]

I'm getting 54429. Wish I knew what I was doing wrong.

[ashmakes]

Would you mind linking the code to your GitHub or in the decription if possible. Thx