The posts here are visible in the homepage for everyone too if not marked as spoilers. That's why when disclosing information about the solution, they should be marked. There are comments that don't need it, see the ones below, most of them don't.

Chrono79, what are you doing in the comments of a solution, if not spoiling the solution? 🤔

You're not alone =)

1. It seems that now all the fans of light mode are satisfied. =)
2. The table is preloaded as int[] - it is convenient, I did the same way in my solution.

Ok nmv I fixed the darkmode lightmode bug

Thanks for solving it man :)

1. I have no idea how to fix that :)
2. The table is preloaded as int[], should I use an ArrayList<Integer> instead?
3. Yes I know :D

In few words(very simplistic):

1 step: we take original GP (l=18) increase it with LT from MP (l=9), then apply XOR with original MP (l=9) and finally have new MP (discarding: l=18-1 = 17)
2 step: we take original GP (l=18) increase it with LT from new MP (l=17), then apply XOR with new MP (l=17) and again have new MP (discarding: l=18-1 = 17)

if after XOR we have 2 zero lead terms, then have to discard them both and jump over next step (!):
3 step: we take original GP (l=18) increase it with LT from new MP (l=17), then apply XOR with new MP (l=17) and again have new MP (discarding: l=18-2 = 16)
4 step: skip
5 step: we take original GP...

GP - generator polynomial, length = 18
LT - leading term of message polynomial
MP - message polynomial, length = 9 at first step, and 17 at every other step

I suffered a little searching for errors, but finally decided. A few suggestions and comments:

1. In the dark theme, the log works as it should, but in the light theme (which I am using) the colors are inverted, so the QRcode picture in log does not look as it should. You have to do something about it.
2. At first I thought that all areas that are not related to the message being encoded should be simply ignored. I was wrong. As a result, I had to create an additional method getBlankQRcode(), which makes a template based on your example.
3. The tables for translation had to be written manually. Although you have written that they are preloaded, but I have not figured out all the possibilities yet =)
4. The link to "Show polynomial division steps" is very valuable. Without it, I would not have figured it out. You described everything well, but this link is a great addition.

Basically, I wasted a lot of time on "Very rarely it can happen that ...." =))

Thanks for the great kata!

It has become smaller than the original generator polynomial.

At second step take lead term from new message polynomial

Good luck :)

Sorry, I have 2 stupid questions =)

First we had message polynomial with length 9.
Then we get through first step (lead term was 64) and our new message polynomial became 17 length. But here is written that "this is our new message polynomial, it has become smaller".

So first question is why it has become smaller if length = 9 >>> length = 17.

The second one is: which term will be the lead term at second step of our example? 168 (from new message polynomial) or 36 (from original MP)?

Thanks!

I like it. All solutions sorting elements first are not efficient, i think.
What's the point of sort all elements first in for example a million elements array if unique element at it's beginning.

Java is available now