# What is the Sandbox?

This "Sandbox" is a place where Code Golf users can get feedback on prospective challenges they wish to post to the main page. This is useful because writing a clear and fully specified challenge on your first try can be difficult, and there is a much better chance of your challenge being well received if you post it in the Sandbox first.

To add an inline tag to a proposal use shortcut link syntax with a prefix: [tag:king-of-the-hill]

## Background

I'm taking AP stats now, so I figured I'd bring the joy to you all here.

The form of an LSRL (least squares regression line) is ŷ = a + bx.

## Input

Your job is to find the least squares regression line for two data sets (lists), X and Y. X and Y will be the same size with at least 2 elements. They will contain positive or negative floats. Input can be taken in any reasonable format.

## Output

The values a and b in the least squares regression line (LSRL) for the data in X and Y, where a is the y-intercept and b is the slope of the LSRL. This page describes a way to calculate this line.

This is , so shortest code in bytes wins!

Also, built-ins are allowed!

Examples

[1.0, 2.0, 3.0, 4.0] [12.0, 15.0, 16.0, 17.0] -> A: 11.0 B: 1.6
[1.5, 1.5, 51.0, -15.0] [2.0, 1109.0, 0.003, 2.0] -> A: 314.92 B: -3.7609
[1.0, 2.0] [1.0, 2.0] -> A: 0.0 B: 1.0
[-15.0, -13.0, -163.0, -12.25, -14.41, 100000.0, -630.0, -135.0] [-142531.0, -12.0, 153.0, -135.0, 135.0, 16.0, 21.0, 0.0] -> A: -20285.0 B: 0.20121
[3.0, 5.0, 7.0, 9.0, 11.0, 13.0] [2.0, 4.0, 6.0, 8.0, 10.0, 12.0] -> A: -1.0 B: 1.0

• 1. Use the full name in the title, because the title will be shown in contexts where the body isn't. 2. What kind of numbers will X and Y contain? The test cases use signed integers, but can answers assume that all inputs will be integers, or that all inputs will be floating point numbers? 3. In some contexts, a quadratic can validly be described as a least squares regression line. It would be better to move the definition which makes it clear that it's a linear polynomial to the introduction, before the section on input, rather than having it in the section on output. – Peter Taylor Oct 22 '16 at 20:55
• 4. There's more than one way to calculate the gradient, and the method described is not the best one. I would suggest removing that section entirely and having a link to en.wikipedia.org/wiki/Simple_linear_regression in the introduction. – Peter Taylor Oct 22 '16 at 20:56
• @PeterTaylor, I have edited the question. Is it better now? – Daniel Oct 22 '16 at 21:26
• Every math language has a builtin for this so it's just solve(x,y). Even without the built-in linear solver, this is just the pseudoinverse of (1,x) multiplied with y. – Angs Oct 23 '16 at 6:04
• I disagree with Peter - for the sake of self-containedness, you should include at least one method of computing the LSRL (and keep the link in case people want to see more methods). – user45941 Oct 24 '16 at 6:50

# 2D Poor Man's Plotting!

## Introduction

Surely you've also come across the problem where you have a multi-dimensional function at hand and want to visualize it (it's just such an everyday problem!). So we need a (programmatic) solution to this. A memorable solution, thus obviously the shortest solution is best!

## Specification

### Input

Your (primary) input will be a function that maps two floats onto another float.

Additionally there are six other parameters (floats), which you may take explicitely or hard-code:

• StartingValueX: 0
• StartingValueY: 0
• EndingValueX: 10
• EndingValueY: 10
• StepSizeX: 1
• StepSizeY: 1

You may take these parameters using your preferred method of input, but please document it (especially how the function should be passed).

### Output

The output is a formatted string or a printed output (whatever pleases you).

### What to do?

Given a function that takes two floats as inputs, iterate it two-dimensionally and print the resulting values. So the top left entry is the result of the evaluation at (StartValueX,StartValueY) and the bottom right entry is the evaluation of (EndValueX,EndValueY) and then you form basically a table from this by incrementing the StartValue by the respective StepSize for each iteration.

Formatting guidelines:
Between each line may be one (or more) empty lines (eg end your lines with \n\n in C).
There must be at least one white space between every output number.
All numbers with the same x-coordinate must be aligned respectively to their start.

### Potential Corner Cases

You may always assume a certain function prototype is used.
If EndValue-StartValue is not an exact multiple of StepSize (which may happen because we're dealing with floats here) then you shall compute for all StartValue + k * StepSize smaller than EndValue as well as for EndValue.
You may always assume the computation results (of the function) are actually representable in your language.
You may not assume that the function is a simple mathematical function (e.g. any valid function adhering to the above conditions in your language must work).
Trailing newlines and white spaces are allowed.

### Who wins?

This is code-golf so the shortest answer in bytes wins!
Standard rules apply of course.

## Test-cases

Default parameters, f(x,y)=x+y
Output:

0  1  2  3  4  5  6  7  8  9  10
1  2  3  4  5  6  7  8  9  10 11
2  3  4  5  6  7  8  9  10 11 12
3  4  5  6  7  8  9  10 11 12 13
4  5  6  7  8  9  10 11 12 13 14
5  6  7  8  9  10 11 12 13 14 15
6  7  8  9  10 11 12 13 14 15 16
7  8  9  10 11 12 13 14 15 16 17
8  9  10 11 12 13 14 15 16 17 18
9  10 11 12 13 14 15 16 17 18 19
10 11 12 13 14 15 16 17 18 19 20


# [Lang1] vs [Lang2] ... vs [LangN]: Battle of the golfers!

and other tags for the challenge questions, which may be a lot.

Note: I haven't figured out all the kinks out yet, all I have is a general idea that, with your help, can be put properly into an awesome challenge (I hope). The parts where my ideas are not solid yet have Notes attached to them that must be read.

# Team

In the midst of the nation of SE, there lies a golf field for the community of golfers. These golfers play day and night using specialised clubs (languages) to complete difficult and challenging holes (challenges). Now it is time to see the best of these golf clubs in the game!

Note 1: First of all, this challenge is going to be language-specific. I haven't decided the languages yet, but right now I'm thinking around the lines of Java, C, C++, C# since these languages are very well known, unlike the other esoteric languages.

You, the golfer, are to choose one golf club out from these [Lang1], [Lang2], ..., [LangN]. You will team up with other players using the same golf club as you.

Note 2: Now there is supposed to be a table where the players can edit into this challenge (once I post it) to join a team. After ~1 week from the beginning of the sign-up, people will start competing in the challenge.

This will be a sort of golf relay, where every player must only complete 1 hole, no more, no less. But of course, each player may help teammates to complete the holes using the minimum number of shots (bytes) as possible. All golfers, of every team, will be playing using the same playing field, in this case UTF-8.

# The Challenge

There will be a list of challenges [see Note 3] to be completed by the team. Each team member must complete 1 challenge only. After a fixed amount of time (see Note 3), all team members must have complete their assigned challenges (see Note 4). I will sum of all the bytes used by each team to figure out the winning team (see Note 5)!

Note 3: Either I will provide all the challenges, or each team member may give 1 challenge to the challenge pool. Challenges can range from anything like quines and "Hello, World" challenges, to more complicated ones, like solving advanced mathematical (pardon me) mumbo jumbo and outputting complex sequences. Every team completes all of these challenges (see Note 4).
The time that I have decided, so far, is to give the entire team 3 weeks (subject to change) to complete their challenges.

Note 4: Each team member complete 1 challenge. However, if the team sizes are different, the team with the greatest number of members will have to complete all of the challenges. The other teams only have to complete as many challenges as members, so that there will be one for every member.

Note 5: The scoring will be the number of bytes used by each team, summed up, and then divided by the number of challenges completed.

Here is the format to each answer:

#Language, challenge-number, byte-count in UTF-8


for example:

#Java, #34, <s>463</s> <s>444</s>  <s>442</s> bytes


## Summary

To sum it up, here is the process that this challenge goes through:

1. People who want to play must choose one team to be on, if they want to play.
2. After registration is over, the games begin: each player is to choose only one challenge and give a golfed answer. (Other players of the same team are allowed to help teammates)
3. After some time, when all the challenges are attempted (one for every person, if people haven't answered in this time, they won't be counted towards the challenge), I will calculate the score of each team (using the total bytes used).
4. Leaderboard is announced and the winning team walks away with all the glory!

Comment any ways this challenge can be improved. My general idea is to implement a team game, instead of the standard free-for-all.

# King of the Hill: Moon Rockets

This is a challenge where you code bots to compete in an arena. In this case, you code a rocket flying over the Moon. Your goal is to shoot down enemy rockets, while not running out of fuel and crashing into the Moon.

### Environment

Unlike many KOTH challenges the playing field is continuous, rather than discrete. You are provided with variables for your position, velocity, and the direction you are facing. Each turn, you are given a list of all the rockets you can "see" (all rockets less than 10 units away). You return a turning speed, thrust and whether you want to fire missiles.

### Arena

The arena wraps left-right. There is a floor; if you crash into it going faster than 4 m/s, you will explode. There is no ceiling, but you do have a limit of 10 fuel units. This can be refueled by landing on the floor going less than 4 m/s.

### Missiles

Every turn, you have the option to fire missiles. Missiles go in the direction you are pointing, at a constant velocity 5 m/s faster than you were when you fired. If they come within a 1m radius of another rocket, they explode and destroy it.

### Gravity

Gravity is a constant force of 1 downward. With a max thrust of 2, you can accelerate at an equal rate in the opposite direction.

Pastebin for current controller code: http://pastebin.com/FNyjrJfr
No tournament code is in place yet.

• Possibly relevant: Spacewar KotH – trichoplax Oct 25 '16 at 20:11
• The controller seems to set the step size at 0.5s. Does that give enough time to take evasive action at all, given the missile speed and sight range? (Even discounting step size, with max thrust of 2, it seems like it's going to be hard to outmaneuver something going possibly several times faster) – Geobits Oct 25 '16 at 20:11
• Additionally related to step size, how are collisions handled? If I'm a meter away from you and fire a missile that goes 7m/s, can it miss by overshooting in that half second? – Geobits Oct 25 '16 at 20:14
• Can spaceships collide with each other, or only the surface? – Geobits Oct 25 '16 at 20:15
• 1. Continuous presumably means floats, which means possible disagreements over the results (because e.g. Java doesn't strictly apply the IEEE-754 spec unless you tell it to). 2. The spec should say at the very least what information is available to the bots. Does the "list of all the rockets you can "see"" include velocity, angular velocity, etc. or just position? – Peter Taylor Oct 25 '16 at 20:24
• I assume one win condition is "last rocket flying", but what about stalemates? Is there a "last turn", since it seems like a couple pacifists could land/refuel forever? – Geobits Oct 25 '16 at 20:27
• @Geobits - the collisions are handled point-by-point, so yes, you could miss someone right in front of you if you have a high closing speed. I think evasion is possible as long as you have at least 1 step of warning. – Skyler Oct 26 '16 at 13:10
• @Geobits Spaceships can't collide in the current code. If they could, what should define a "collision"? Currently they are modeled as point objects. – Skyler Oct 26 '16 at 13:11
• @PeterTaylor they can access velocity etc. as well as position. I think this might be required as if a rocket is moving fast you need to lead it with the missile. – Skyler Oct 26 '16 at 13:12
• @Geobits I could add a condition that limits the total turns. Does 50 turns seem reasonable? – Skyler Oct 26 '16 at 13:13

# C Compiler

Implement a basic C Compiler.

You do not need to implement the libraries. Compiler would output an object file containing a function, that would be linked with some libraries using GNU binutils ld and run.

No function calls, unions, structs, casts will be used in the test cases.

Test case:

int test(int i, int j) {
int x[10][20];
int s = i + j;
i = 0;
j = 0;
for(i = 0; i < 10; i++)
while (j < 20)
{
j += 1;
x[i][j] = i * j;
}
for(i = 0; i < 10; i++)
while (j < 20) {
s += x[i][j];
}
return s;
}


• What does need to be implemented? A question should ideally be self-contained, but if there's too much information to fit in the post then it should contain links to the relevant specifications. – Peter Taylor Oct 26 '16 at 10:08

# HamSort a list of strings

Given a list of strings, sort them in a manner such that the Hamming distance between each is minimal.

The Hamming distance between two strings of equal length is the number of positions where the corresponding characters do not match. For example, the Hamming distance between apple and ample is 1 since only the second characters of each do not match.

apple
ample
*


# Rules

• This is so the shortest code wins.
• Each string in the input list will have the same length and will consist of only the letters in the English alphabet. You may assume all to be either uppercase or lowercase.
• If there are multiple possible solutions, you may output any number of them.

## Test Cases

<to be created>

• The Hamming between consecutive pairs? If so, what is to be minimized? The sum of the Hamming distances? Their maximum? – xnor Oct 26 '16 at 7:32
• What is (for this challenge) the hamming distance of strings without equal length? Or will the list contain only strings of the same length? – Laikoni Oct 26 '16 at 8:06
• @xnor It's still a work-in-progress. I'd like to have the problem be minimizing the Hamming distance between the consecutive pairs, but that problem might be hard enough that the only golfy solution would be generating all permutations and testing each pair. Then a length n input would have a length n-1 Hamming distance array where each value is minimal. If you have better ideas that could still capture the spirit of this, feel free to suggest them. – miles Oct 26 '16 at 8:33
• @Laikoni Yes, Hamming distance is only defined between strings with equal length and one of the rules was that the input array would only contain strings with equal length. – miles Oct 26 '16 at 8:34
• I think it's NP hard to find the path that minimizes total Hamming distance by reduction to Ham-Path. So, I wouldn't expect anything better than brute force there. – xnor Oct 26 '16 at 9:00

How fast (slow) is your language?

This is a question both to test your coding skills but also to speed benchmark your favorite language. The code snippet will make a table of the fastest code per language. It is related to previous [challenge]1 so you may find some of the code there a useful starting point.

The challenge is to write the fastest code possible for computing the permanent of a matrix whose entries are floating point in the range -1 to 1.

The permanent of an n-by-n matrix A = (ai,j) is defined as

Here S_n represents the set of all permutations of [1, n].

As an example (from the wiki):

In this question matrices are all square and will only have the floating point values in the range -1 to 1 (excluding both ends). I will construct all tests in such a way that the permanent is also in the range -1 to 1.

Examples

Input:

[[-0.34389946 -0.811193  ]
[ 0.62778788 -0.09918764]]


Permanent:

-0.47514655471


Input:

[[-0.34389946 -0.811193  ]
[ 0.4783581  -0.56636227]]


Permanent:

-0.19326906099


You should write code that, given an n by n matrix, outputs its permanent.

As the output will never be bigger than 1 in absolute value you do not have to worry about overflow. However you do have to worry about numerical accuracy. All answers should be correct up to 8 decimal places.

Languages and libraries

You can use any available language and libraries you like but no pre-existing function to compute the permanent.

More to be added..Work in progress

• "All answers should be correct up to 8 decimal places." Are you sure this is possible? I suspect that finding the permanent has similar problems of ill-conditioning to finding the determinant. – Peter Taylor Oct 28 '16 at 11:37
• So this is your posted permanent challenge but with floats? I don't see what that's so different as to warrant another challenge. – xnor Oct 29 '16 at 6:12
• @xnor The only differences are a) it avoids all the big int stuff that turned out to be crucial in the last challenge and b) it will have a code snippet and will be designed to encourage answers in lots of languages. Not just fast ones. The idea is to show how slow (fast) your language is and it can serve as a fun sort of language competition/benchmark. – user9206 Oct 29 '16 at 7:07
• @PeterTaylor That's a very interesting point. I am not sure of the answer but I was going to choose random orthogonal matrices (I need to change and add examples) so maybe ill-conditioning will be very rare? – user9206 Oct 29 '16 at 7:08

# Directional Brainf*ck Compiler/Interpreter code-golf

Challenge: Write a Brainf*ck compiler with 4 extra commands.

L Set the direction to go left
R Set the direction to go right
D Set the direction to go down
U Set the direction to go up


The original commands (altered):

+ Increment the cell one
- Decrement the cell one
[ Start a loop, if the cell is 0 then go to the next ]
] end a loop if the current cell is 0. Else go back to the last [ read and continue from there (in the direction you were going last time it was read.
. output the byte at the data pointer.
, accept one byte of input, storing its value in the byte at the data pointer.
> point to the cell to the right
< point to the cell to the left


## Compiler/Interpreter

In this case the compiler must either convert the code to another language or be an interpreter and run the code.

This is basically a directional version of brainf*ck. Directions should work the same way they do in ><>.

The pointer should start at the tape's beginning. It must be infinite in the right direction. Left infinite is optional (if you add it -10 to the score). Upon end of code output (in any reasonable form) the current values on all the stacks. If a ] is hit before a [ output a warning (with a message of your choosing). If a , is met at the end of the code you can do with it as you please. The program will terminate at the end of the file.

Standard loopholes apply.

• So, what does ] do – feersum Oct 23 '16 at 19:05
• Taken from Wikipedia: if the byte at the data pointer is nonzero, then instead of moving the instruction pointer forward to the next command, jump it back to the command after the matching [ command. @feersum – XiKuuKy Oct 24 '16 at 0:51
• You don't need to tell me what it does in brainfuck. I'm asking you, what does it do in this language you're defining here? – feersum Oct 24 '16 at 1:02
• @feersum Oh okay. If the byte at the data pointer isn't 0 then instead of going to the next command. Go back to the last [ found. Should I add that to the challenge description? – XiKuuKy Oct 24 '16 at 1:39
• That description doesn't accord with the normal interpretation, which goes to a matching bracket, rather than the last one found. And it still doesn't clarify how it will interact with the two-dimensional aspect. In particular, what direction will the instruction pointer travel after jumping to a [? – feersum Oct 24 '16 at 2:01
• @feersum it should go in the same direction it was going last time it was parsed I guess. I thought the last [ parsed would be better since I can't imagine matching brackets in 2D language. – XiKuuKy Oct 24 '16 at 2:06
• You need to specify the behavior in the challenge specification. Personally I don't see an issue with jumping back to the command after the last [ encountered, continuing in the same direction. 1D programs in this extension that didn't use any of LRDU would be compatible with regular brainfuck with that behavior (assuming no mismatched braces). – user45941 Oct 24 '16 at 6:47
• Also, what does [ do if the cell is 0.... – feersum Oct 24 '16 at 19:03
• @feersum Jump to the next ] in the current direction, presumably. That needs to be specified too, though. – user45941 Oct 24 '16 at 21:09
• @trichoplax Done. – XiKuuKy Oct 25 '16 at 3:43
• Since tags can't be added to an answer, you can use [tag:code-golf] to show the tags you are going to use, so it won't be necessary in the title. – trichoplax Oct 25 '16 at 9:17
• If it's going to be a compiler/interpreter, it may attract more competitors if that is included in the title. – trichoplax Oct 25 '16 at 9:18
• Is this ready to post? – XiKuuKy Oct 28 '16 at 14:36
• It's missing a lot of info. (1) The position and direction of instrucitons pointer at the beginning. (2) What happens upon reaching an edge of teh code. (3) What happens if ] is hit before any [. (4) The usual BF-specific stuff: (a) behavior of , on EOF (b) Integer range of cells (c) whether tape should be infinite in one direction, both, or something else. (5) How does the program terminate, if at all. – feersum Oct 29 '16 at 2:52

# Modular Reverse Recursion Sequence

Given any number n > 0, let r represent the reverse of the number n. Iterate until the final result is zero, passing the result of each iteration back into the function using recursion or a methodology of your choice by performing the below operation:

• If r > n for that iteration the result is r % n.
• If n > r for that iteration the result is n % r.
• If n % r = 0 or r % n = 0, you terminate iteration.

Take the intermediate result of each execution and store them in an array for the final answer.

Lets walk through an example where n=32452345.

54325423 % 32452345 = 21873078 # r > n, uses r % n
87037812 % 21873078 = 21418578 # r > n, uses r % n
87581412 % 21418578 = 1907100  # r > n, uses r % n
1907100 % 17091 = 9999         # n > r, uses n % r
9999 % 9999 = 0                # r % n = n % r = 0, terminated

Result: [21873078, 21418578, 1907100, 9999]


Another example n=12345678:

87654321 % 12345678 = 1234575 # r > n, uses r % n
5754321 % 1234575 = 816021    # r > n, uses r % n
816021 % 120618 = 92313       # n > r, uses n % r
92313 % 31329 = 29655         # n > r, uses n % r
55692 % 29655 = 26037         # r > n, uses r % n
73062 % 26037 = 20988         # r > n, uses r % n
88902 % 20988 = 4950          # r > n, uses r % n
4950 % 594 = 198              # n > r, uses n % r
891 % 198 = 99                # r > n, uses r % n
99 % 99 = 0                   # r % n = n % r = 0, terminated

Result: [1234575, 816021, 92313, 29655, 26037, 20988, 4950, 198, 99]


A final example n=11000:

11000 % 11 = 0 # n % r = 0, terminated

Result: []

• If you have to include 0 as the final element, it might be better to say "if n = 0, terminate", since terminating when n=r doesn't really include 0, because the final modulus (e.g. 9999%9999) was never done. – Geobits Nov 1 '16 at 18:22
• @Geobits Edited to remove 0 from the list of final results, nice catch. That was the only part I did manually. – Magic Octopus Urn Nov 1 '16 at 18:33

### Simplified maze-solver

A spin-off from this question Textual maze solver

Write a program, that will take an bunch of 0's and 1's as it's input (see below), and a half-solved maze as it's output. The program should take 1's as walls and 0's as corridors. The output should should show 99's as walls and a number as the distance from the start of the maze. The maze should have no real entrance/exit; instead, the entrance is the square in the 2nd column of the 2nd row.

Examples (allowed cases):

Input               Output

1 1 1 1 1 1         99 99 99 99 99 99
1 0 0 1 0 1         99 00 01 99 05 99
1 1 0 0 0 1   -->   99 99 02 03 04 99
1 0 1 1 0 1         99 10 99 99 05 99
1 0 0 0 0 1         99 09 08 07 06 99
1 1 1 1 1 1         99 99 99 99 99 99

{{1, 1, 1, 1, 1, 1, 1, 1, 1},         99 99 99 99 99 99 99 99 99
{1, 0, 0, 0, 1, 0, 0, 0, 1},         99 0  1  2  99 16 17 18 99
{1, 1, 1, 0, 1, 0, 1, 0, 1},         99 99 99 3  99 15 99 19 99
{1, 0, 0, 0, 1, 0, 1, 0, 1},   -->   99 6  5  4  99 14 99 20 99
{1, 0, 1, 1, 1, 0, 1, 0, 1},         99 7  99 99 99 13 99 21 99
{1, 0, 0, 0, 0, 0, 1, 0, 1},         99 8  9  10 11 12 99 22 99
{1, 1, 1, 1, 1, 1, 1, 1, 1}}         99 99 99 99 99 99 99 99 99


Examples (not allowed outputs and inputs):

11111 10101 10101 10001 11111  --> 9999999999 9900990699 9901990599 9902030499 9999999999


Notes:

• as seen above, the input and output must be bunch of numbers separated by whitespaces, commas, full stops, newlines, parenthesis (and so on), and mustn't be in a single row

• you can output 01 instead of 1

• cases where the distance is over 99 don't need to be covered

• the program doesn't need to solve the maze, just to find the distance from the start

• two fields connected diagonally aren't next to each other

• it's , so the shortest answer in bytes wins!

• it makes more sense to show walls as ## or || etc, instead of an integer – FlipTack Nov 5 '16 at 21:56
• The restrictions on the format on the output and inputs don't add anything to the challenge (imo) : the interesting part is solving the maze. – Dada Nov 6 '16 at 10:56
• @Dada They allow the program to be shorter, at least in most languages I know. – RudolfJelin Nov 6 '16 at 13:55
• @Flp.Tkc See the comment above. – RudolfJelin Nov 6 '16 at 13:56

Defs need to fix up layout

[Very much a WIP]

# Maze battle

In this challenge, bots will team up, navigate around a maze and try to destroy the enemy team's artifact, and protect their own.

A maze might look like this. Note that the bots will receive numbers rather than ascii art, which represent their vision, and the objects they see.

+---------+-+---+-------+++-+-+---+-+-------+-+-----+
|         | |   |    #  | | | |   | |       | |     |
+-+ +-+-+ + + + +-+-+ +-+ + + + + + + + +-+-+ +-+ + |
|   |         |         |     | |   | | |   |     | |
| +-+-+ + +-+ +-+ + + +-+ + +-+ +-+ +-+ + +-+ +-+ + |
|       | |   |   | |     | |   |         | |   |   |
| +-+-+-+ +-+-+-+ + + + + +-+ +-+-+ +-+ + + + +-+ + |
|   |     |       |   | |   |   |     | | |       | |
| + + +-+-+ +-+ +-+ +-+ +-+ + +-+ + + + + + +-+ + + |
| |       |                 | |   | | |     |   | | |
| + + + +-+-+-+-+ +-+-+-+ + + +-+ +-+ + +-+-+-+ +-+-+
| | | |     |   | | |     |     |   | | |           |
| +-+ +-+ + +-+ +-+ +-+-+ +-+ + + + + + +-+-+ +-+ + |
|   | |   | |           | |   | | | |   | | |     | |
| +-+ + +-+ +-+ + + + +-+-+-+ +-+ +-+-+ + + + +-+ + |
|     | |       | | |   | | | | | |         |   | | |
| + + +-+ + + + + +-+ +-+ + +-+ + +-+ +-+-+ +-+ +-+-+
| | |   | | | | |   |   |       | |   | |       |   |
| +-+ + + +-+ +-+-+-+ +-+ +-+-+ + + +-+ + +-+-+ +-+ |
|   | |   |         |       |   |       |   |       |
+-+ +-+-+ +-+ +-+ +-+ + +-+ + + + +-+ +-+ +-+ + +-+-+
|   |   |   |   |   | | |     |     |   | |   |     |
| + +-+ +-+ + +-+-+-+ + + + +-+-+-+ +-+ + + +-+ +-+ |
| |       |   |   |     | |         | |   | |       |
| +-+-+ + +-+ +-+ + +-+-+-+ + + + +-+ +-+ +-+-+ +-+ |
| |   | | | |       | |   | | | | |   |     |     | |
+-+-+ + +-+ +-+-+-+-+ +-+ +-+ +-+ + +-+-+ +-+ +-+ + |
| | | | |   |           |     | |     |       | | | |
| + + +-+-+ + +-+ +-+-+ + +-+-+ +-+ +-+-+-+ +-+ +-+-+
|   |     | | |             |   | |   | |     |     |
+-+ +-+-+ + + + +-+ +-+-+-+ +-+ + +-+ + +-+ +-+ +-+-+
|   | |   |   |   | |     | |   |     |   |       | |
| +-+ + +-+-+ + +-+ + + +-+-+ + + + + + +-+-+ + + + |
|       | |           |       |   | | | |     | |   |
| +-+-+ + +-+ + +-+-+ + + +-+ + +-+-+ + +-+ + + +-+ |
|   | |   |         | | | |   |     |   | | | | |   |
+-+ + + + + + + +-+ + +-+-+-+-+ +-+-+ +-+ +-+-+ + +-+
|       |   | | |           |     | | |   | |       |
+-+-+-+ + +-+-+-+-+-+ + +-+-+ +-+ + +-+ + + + +-+ + |
|     |   | |         |     | |     |   |       | | |
+-+ +-+ +-+ +-+ + +-+ +-+-+ + + +-+ +-+ +-+ +-+ +-+-+
|     | | |     |   |         | |           |   |   |
+-+ +-+-+ + +-+-+ +-+-+ +-+-+ +-+ +-+-+-+-+ + +-+ + |
|               | |   |                 |   | |   | |
+-+-+ +-+-+-+ + + + + +-+-+-+-+-+-+ +-+ + + +-+ +-+ |
|       |   | | |   |   |           | | | |       | |
| +-+-+ + + +-+ +-+ + + +-+-+ +-+ + + + +-+-+-+-+ + |
| |       | | | |   | |       |   |     |           |
+-+-+ + +-+-+ + + +-+ +-+ +-+-+ + +-+ + + +-+-+ +-+-+
| |   |     |   | | |   |       |     | |     |     |
| + +-+ + + + + + + + + + + + + +-+-+-+ +-+-+ +-+-+ |
|       | | | |    #  |   | | | |       |     |     |
+-------+-+-+-+-------+---+-+-+-+-------+-----+-----+


Bots receive information about their current environment. They will receive a 8 lines of numbers, representing their view in each of the directions. vision is blocked by walls, and walls only, so at the end of the line of sight is a wall.

Numbers represent the following things:

• 0 represents a space with nothing in it.
• 1 represents a friendly bot
• 2 represents an enemy bot
• 3 represents your team's artifact
• 4 represents the enemy team's artifact

To destroy the enemy's artifact, you have to shoot it a bunch. heaps. So, you probably want to bring that artifact to your team, or your team's side, so that you can shoot it more.

When you die, you respawn after 5 turns.

At the start of the round, bots spawn into the first third of the maze on their side. when respawning, they also spawn into this area.

• Nice idea! Would be better if it was Java :D – user41805 Nov 6 '16 at 10:20
• @KritixiLithos: No – Destructible Lemon Nov 8 '16 at 23:28

# Format an XML file

Your task is to parse an XML snippet (not necessarily an entire file) and format it. I am explicitly making this simple to encourage extremely golfed code.

The formatter should transform XML like this:

<?xml version="1.0" encoding="UTF-8"?>
<body>
<div class="community-option">
<input id="communitymode" name="communitymode" type="checkbox"/>
<label for="communitymode" title="Marking an answer community wiki encourages others to edit it by lowering the reputation barrier required to edit. However, you will not gain any upvote reputation from it. This cannot be undone.">community wiki</label>
<a href="www.hoehel.be"><i>My inner web developer hates me</i></a>
<b>BOLD</b>
</div>
<iframe/>
</body>


into formatted text like this:

?xml
body
div
input
label
a
i
b
iframe


## Rules

The main element does not get a indent, every layer deeper does get an extra indent. An indent is either a tab or 4 spaces.

When there are multiple root elements, they simply are added below eachother with no indentation. The ?xml element is not different from any other element.

The XML will have a space after each element name, so if there is a CDATA element, it will be formatted as <![CDATA[ (the data)]]> and will need to be formatted as ![CDATA[ at the correct indentation.

Also, the XML will not include > and < except for opening and closing elements.

• In XML, a right angle bracket > is allowed in an attribute. So perhaps include that as a test case, or explicitly state that this will not happen. Also, what qualifies as an indent? One space? Four spaces? A tab character \t? – Sanchises Nov 8 '16 at 11:31
• 1. The example is invalid because (a) it doesn't have a single root element; (b) the input element isn't closed. 2. This isn't really pretty printing XML. It's pretty printing the element structure of XML. 3. There are a whole bunch of important test cases which are missing. CDATA. Namespaces defined at the root element. Namespaces defined on non-root elements. Elements expanded from entities. There are bound to be some others if you look at the spec. – Peter Taylor Nov 8 '16 at 17:08
• @sanchises Reworked the question to avoid problems there. – vrwim Nov 8 '16 at 21:11
• @PeterTaylor (a) I'll include that it can be a snippet. (b) Didn't notice it, took some HTML from stackoverflow source code, don't blame me :) 2. Renamed to "formatting" the XML 3. Added workaround for the special cases I could find. – vrwim Nov 8 '16 at 21:15

# Build a learning 20Q bot

Twenty Questions is a verbal game where an 'asker' asks you a series of yes or no questions about an object that you're thinking of. If the 'asker' gets the object correct after twenty questions, then the 'asker' wins the game. If the asker does not guess the question right after 20 questions, then the 'thinker' wins.

## Input:

Your program needs to take 4 different inputs, non simultaneously. Those three inputs are:

• A category (valid categories are Animal, Food, Action, Person/Character, Thing)
• A yes / no input from the user to answer the questions.
• A question input to 'teach the bot', and an object input to show the bot what was being described.

## Output:

Your program needs to play a game of 20 questions, wherein it follows this structure:

• Ask a category like so: "Select a category (Animal, Food, Action, Person/Character, Thing):"
• Ask questions based on the category.
• Make a guess of the object like so: "Is it a "[guess]"?".
• IF the guess was wrong, ask what the object was, and inquire the difference between the guess and the object described.
• your program must stop at 20 questions and make a guess. it doesn't have to be random.
• your program must progressively 'learn' as it is played. It must be able to add new questions and objects to guess.

## Examples:

⇨Select a category (Animal, Food, Action, Person/Character, Thing):
⇦Food
⇨Is it  ?
⇦No
⇨What is it?
⇦Marmite
⇨What is the difference between  and Marmite?
⇦It has a saying 'You either love it or hate it'
⇨
⇨Select a category (Animal, Food, Action, Person/Character, Thing):
⇦Food
⇨Does It has a saying 'You either love it or hate it'?
⇦No
⇨Is it Marmite?
⇦No
⇨What is it?
⇦A Carrot
⇨What is the difference between Marmite and A Carrot?
⇦It has lots of vitamin C
⇨
⇨Select a category (Animal, Food, Action, Person/Character, Thing):
⇦Food
⇨Does It has a saying 'You either love it or hate it'?
⇦No
⇨Does it It has lots of vitamin C?
⇦Yes
⇨Is it a Carrot?
⇦No
⇨What is it?
⇦A Yam
⇨ What's the difference between a Carrot and A Yam?
⇦It grows underground.
⇨
⇨Select a category (Animal, Food, Action, Person/Character, Thing):
⇦Food
⇨Does It has a saying 'You either love it or hate it'?
Yes
⇨Is it Marmite?
Yes
⇨Yay, I Win!
⇨
⇨Select a category (Animal, Food, Action, Person/Character, Thing):


Etc.

• Pass2: How Can I improve this? I would like some critiques before I post. – tuskiomi Nov 16 '16 at 19:51
• Possible duplicate: codegolf.stackexchange.com/q/57171/32686 – Blue Feb 10 '17 at 15:14

# Produce a quine of length n

Write a program that, given a number n, produces a quine in your language of length n. If your language cannot have a quine of that length, it does not need to be generated.

Standard loopholes are forbidden.

The quines that are generated must be proper quines.

Remember, this is , so the code with the smallest number of bytes wins.

• Is there a minimum/maximum n? Plenty of languages don't have a 1 or 2 length quine, for example. – Geobits Nov 17 '16 at 20:28
• @Geobits How botu now? – Oliver Ni Nov 18 '16 at 0:03
• Seems reasonable. I'm not sure how to prove it for some languages, but I'll leave that to answerers :) – Geobits Nov 18 '16 at 0:13
• The ! in the title made me think that the quine had to have length n factorial, whereas the text says n. You should remove the ! in the title. – Fatalize Nov 18 '16 at 9:11
• @Fatalize Okay. Done. – Oliver Ni Nov 18 '16 at 16:09
• Regarding what @Geobits said, with the current wording I'm definitely worried since answerers would need to prove that they've got the minimum length quine covered. – Sp3000 Nov 19 '16 at 5:54
• Sp3000's concern is definitely valid. It would mean that you can only answer in languages where the shortest known quine has been proven to be optimal. Also, I'm not sure it's terribly interesting in most languages, because you can usually just insert spaces or no-ops into one or two places and the code remains a quine. – Martin Ender Nov 21 '16 at 15:23

# Rearrange the Words!

Given a string of text containing some words, which are separated by spaces, determine the best arrangement to fit it in the shortest number of pages.

If a word is shorter than the length of a page:

• If the word fits on the latest page, it stays on that page
• If the word does not fit, it goes to the next page.

If a word is longer than the length of a page, it stays on that page and overflows to the next page.

Words have one character between them if they are on the same page.

For example, if the length of the page was 10, and my words are hello my name is oliver ni!!, it would fit like this:

hello my
name is
oliver
ni!!       # 4 pages


However, one of the ways to rearrange it can get a shorter number of pages:

hello name
oliver is
my ni!!    # 3 pages


So the shortest way in this case would be 3 pages.

Remember, this is , so the code with the fewest bytes wins.

• 1) Does the greedy algorithm give an optimal solution? If not, there should be test cases specifically for it. 2) A few test cases would be good, covering edge cases such as everything fitting on one line 3) Can the length of the page be shorter than the longest word? Is that valid input? 4) Is the output the arrangement or the number of pages? – Sp3000 Nov 21 '16 at 1:15
• This problem is well-known to be NP-complete. As such, the greedy algorithm can't possibly be optimal. I imagine that code-golf solutions will just brute-force the solution, given that (for an NP-complete problem) it's very hard to objectively enforce a faster-than-brute-force algorithm. – user62131 Nov 21 '16 at 1:19
• An example where the greedy algorithm fails: word lengths 10, 6, 5, 3, 2, 2, page length 14 (this assumes that spaces have already been "factored in to" the word and page lengths by adding 1 to each). The greedy algorithm gives (10, 3), (6, 5, 2), (2). The best answer is (10, 2, 2), (6, 5, 3). – user62131 Nov 21 '16 at 2:04
• @ais523 I'm pretty sure the greedy algorithm will give an optimal solution. – Oliver Ni Nov 21 '16 at 5:28
• @Oliver ais523's example is actually 9, 5, 4, 2, 1, 1 with page length 13. The greedy algorithm gives (9 + space + 2), (5 + space + 4 + space + 1), (1), but the optimal solution is (9 + space + 1 + space + 1), (5 + space + 4 + space + 2). – ETHproductions Nov 21 '16 at 15:45

Change the type of a variable without introducing any new identifiers while preserving a value resulting from the conversion of the original one

Consider the variable which type you should change is declared/defined inside a function. Another thing to consider is that there wouldn't be any other identifiers manually defined/included before your function and the variable inside it (this includes referring to standard header/modules too). Also you aren't allowed to change the value or meaning of any existing identifiers (if any) except a after your code is finished, compared to before it started executing. (Of-course this doesn't forbids you to actually change them but after your code is finished they must be in the same state as before it was entered in, except a of-course)

As identifiers I mean reserved names of functions, variables and etc.

Example in C:

void f() //no identifiers defined/included prior
//the function can't have any parameters

{ //inside function block

int a = 67; //assign some constant value here

//no new identifiers introduced
//and no existing one have changed meaning (or value) after this point
//except 'a' of-course

a ; //here 'a' must have different static type
//with a value resulting from the conversion
//of the one stored previously in 'a'

//You are allowed to include libraries/modules (identifiers) here in
//order to show us the value and type of 'a'
//The unused statement above is used only as valid syntax containing 'a'
}


Of-course the above snippet is just a sample. In case of using actually C you are allowed to add all other kind of stuff (like additional block scopes, etc). The only important thing to target is that after you have declared some variable in your function you have to write some magic with the requirements given.

This question have intended solution in C. I'm not sure how applicable is it for other languages though because it depends on the concept of having some names (identifiers) referring to certain entities. Also it needs types and object values defined too.

The initial and target type of a aren't specified. Also if the identifier a of the variable is not allowed, you can use one that is.

• I think I understand now your purpose better. This is a programming-puzzle, and we certainly allow them. For that purpose, I think its a good idea to limit it to C. That said, you should be very explicit about where they are allowed to put code and what code they can write. (You have it in your comments, but I'd add it in the post) – Nathan Merrill Nov 21 '16 at 16:33
• @NathanMerrill Thanks for the feedback! Yeah - you are probably right to limit this question as C only. – AnArrayOfFunctions Nov 21 '16 at 16:49
• Is any C standard allowed? E.g. I have a solution which only works with C99. – Peter Taylor Nov 21 '16 at 22:05

# Pluralize Words code-golf

(DISCLAIMER: In this challenge, several exceptions and rules are not accounted for. Many plurals are based on sound and not spelling, and still others have no set guidelines. Please do not use this as an actual set of rules for making plurals.)

A lot of questions here require using different Algebras. ... Algebra? Algebrae?

OK, that's it. We're going to get to the bottom of this.

Your challenge is to write a program that accepts a word as input and outputs the plural (in standard English rules).

## Rules on making plurals

• Words ending in s, x, z, ch, sh, o, ex, or ix have es added to them. (For example, box -> boxes)
• Words ending in y preceded by a consonant have the y removed and ies added to the end. For example, cherry -> cherries. (Words like day with a vowel before the y are treated as normal.)
• Words ending in f or fe are treated as normal.
• Words ending in us have the ending changed to i. (For example, alumnus -> alumni)
• Words ending in is have the ending changed to es. (For example, oasis -> oases)
• Words ending in um are changed to a. (For example, medium -> media)
• Words ending in a are changed to ae. (For example, antenna -> antennae)
• All other words have s added to the end.

Several rules have been omitted for simplicity:

• You do not have to account for irregular plurals like sheep. If sheep was inputted, you would return sheeps.
• The o/f/fe/ex/ix rules have lots of exceptions. knives, matrices, etc. I can't find any pattern, however, so these are simplified to one rule.
• chateau -> chateaux, man -> men, child -> children, foot -> feet, and mouse -> mice have been left out due to ambiguity. You can replace these with chateaus, mans, childs, foots, and mouses, respectively.
• Words are counted based on spelling, not sound. TimmyD mentioned in the Sandbox:

For example, Czech (as in, a person from the Czech Republic) goes to Czechs ("checks") since the ch is pronounced like ck.

Here you would output Czeches (as it is spelled with a ch).

## Input

A single word, like challenge. The input will only contain letters a-z.

## Output

The pluralized word, based on the rules above, like challenges. You may have a trailing newline.

Output may be case-insensitive.

## Test Cases

algebra
algebrae

shoe
shoes

fax
faxes

datum
data

bunny
bunnies

qlwlsh
qlwlshes


# Meta Questions

UPDATE: I've answered these questions myself. Do you guys think it's ready for posting?

• Is this a duplicate? (No, other similar challenges have different scoring)
• Are there any special cases I left out? (Probably but they'd over-complicate things)
• When researching this I discovered there are a LOT of little rules. I've omitted some here, but is this too complex to be a feasible challenge? (No)
• In addition, have I left out too many rules? (Adding any more would make it too complex)
• The problem is that English handles its pluralization (and a/an distinction) based on the vocal pronunciation, and not by the letters themselves. For example, Czech (as in, a person from the Czech Republic) goes to Czechs ("checks") since the ch is pronounced like ck. I think the simplifications you have here are OK, so long as you're very explicit that there are a lot of exceptions that aren't accounted for. – AdmBorkBork Aug 5 '16 at 15:13
• Also, relevant chat discussion – AdmBorkBork Aug 5 '16 at 15:21
• @TimmyD Interesting. I'll double-check the edits I made to make sure those strange cases don't slip through. – ASCIIThenANSI Aug 5 '16 at 15:26
• I think the f rule could be left in, since it's pretty consistent that f and fe become ves. I can only think of a few exceptions to that rule. A better title might be "Pluralize Words" (since pluralize is actually a word). I think if you just need to follow the rules and not worry about exceptions, it wouldn't be too complex. – Business Cat Aug 5 '16 at 15:30
• Also possible duplicate of this and this. However, I prefer this one to the others since this explicitly states the rules your program should follow. – Business Cat Aug 5 '16 at 15:31
• @BusinessCat Those are different, scoring is based on the number of wrong words there. Here, this is a code-golf for shortest code that properly prints plurals (with some exceptions.) – ASCIIThenANSI Aug 5 '16 at 15:47
• My answer to "Is this too complex to be a feasible challenge?" is "Yes". Natural language processing in general (and especially where the language is English) is a better fit for challenges which score by performance against a large suite of manually curated test cases than pure code-golf. – Peter Taylor Nov 23 '16 at 15:54

# Find the Mersenne Twister seed fastest-code

Given a sequence of N 32-bit unsigned integers, find a 32-bit seed for the Mersenne Twister PRNG (specifically, MT19337) such that the first N 32-bit unsigned integers produced are the input sequence.

## Rules

• Solutions will be tested and timed on a free Cloud9 workspace (512 MB RAM, 1 8-core CPU). Full CPU specifications can be found in this Gist.
• There will be 50 sequences used to time solutions, each containing at least 20 and at most 50 integers. All sequences will be generated using CPython 3.4's MT19337 implementation (the source code can be found here).
• Brute force (by way of iterating through all possible seeds) is not allowed, as it would take an obscene amount of time to run all 50 timing cases with such a solution.
• Hard-coding seeds is not allowed, as it undermines the point of this challenge.
• I reserve the right to change the timing sequences to different sequences of similar complexity if needed, such as in the event that two seeds are found for the same sequence.
• Please include instructions on how to run your solution, including installation instructions for any software that is not included with a standard Ubuntu installation.

## Test Cases

Note that the seeds given may not be the only seeds which produce the given sequence (if other seeds are found, please let me know). These smaller test cases are provided for verification purposes only, and will not be used for scoring.

[859552199, 418371998] -> 765976290
[3575669016, 3345750943, 1425199743, 3641622143, 468413882, 1201820413, 440223465, 4032367274, 1303769163] -> 3089232394
[3301913871, 2465802243, 3612732284, 2327984155, 3125193467, 1186355054, 3586515345, 1676699096] -> 3317482306
[1187719129, 4252341260, 3652285377, 2546267472, 259146681, 3132024105] -> 618872552
[567043356, 3386297553, 3291980436] -> 2064328469
[2371812852] -> 3402334589
[2169144927, 1612138200, 2367065602, 3106703576, 3360699020, 2044583150] -> 1508970517
[4264152408, 93510901, 2140553491, 2298459244, 3988997159] -> 3805008435
[815695699, 3390998580, 1837233431, 35472600] -> 1298560578
[2785297506, 692193253, 1505379580, 2810620241, 2299478120, 233996693] -> 1856467397


## Timing Cases

These are the 50 sequences, along with the respective seeds that produced them, that will be used for scoring solutions. Once again, if any other seeds are found that produce the given sequence, please let me know.

[3626764237, 1654615998, 3255389356, 3823568514, 1806341205, 173879092, 1112038970, 4146640122, 2195908194, 2087043557, 1739178872, 3943786419, 3366389305, 3564191072, 1302718217, 4156669319, 2046968324, 1537810351, 2505606783, 3829653368, 3900315155, 938204377, 2167613558, 598176026, 1210484339, 600203567, 3246154361, 407295012, 2655874620, 3433407905, 1075916535, 4240245955, 3907330410, 2287343378, 4221031477, 3028691745, 3479856542, 2585149702, 3874773259, 631194409, 1332073689, 424185324, 3134603515, 316721330] -> 0
[3393124633, 2450474977, 1966463085, 3673681320, 1729202683, 1167867720, 325264180, 3917283871, 100160351, 791524511, 2930202830, 3113143968, 1502082504, 1809749915, 418618347, 1516663034, 578608165, 3127414055, 3472262548, 308161377, 135281507, 4131277967, 3953099834, 1042116387, 1985078128, 1789355090, 2975834429, 2756451880, 1853842973] -> 80552670
[3501423986, 1904817736, 3722254077, 4209805889, 624241735, 3072497443, 3348948634, 948724253, 2775088165, 653145320, 699931897, 2271688288, 965327734, 3611254581, 4021716510, 3489649715, 590896686, 2492882151, 2449284001, 1709766651, 2373030892, 1906330141, 3281376004, 542155191, 654946082, 3866707349, 675539429, 3153689518, 2192976919, 2369760520, 1789539597] -> 123863960
[161627675, 758641073, 3882484966, 1142624247, 4128058828, 2063104815, 1532506610, 4133456839, 2543581728, 18178196, 2276191025, 6278001, 3128406689, 2146406812, 39141315, 1050571803, 2511216100, 2411368419, 3269811767, 815334548, 3812711899] -> 178693462
[1006603821, 2040840228, 2217261808, 3301867033, 3944952417, 1692780898, 40334807, 2573488161, 1195239502, 1063781197, 758636410, 1944506134, 1022411961, 3709537934, 175021745, 3183030828, 1196557078, 11946672, 1551403327, 4143434962, 3283348277, 1825370082, 2293735756, 429108850, 745523037] -> 227497455
[2252079699, 1086590868, 525211993, 209343944, 1861419840, 2057307759, 2083384643, 2590024292, 4135416707, 3508386594, 1358765199, 3089076939, 3719484781, 516962076, 4107295833, 723033883, 1857756581, 434940583, 919038598, 2441529245, 1687894943, 13982251, 1940221612, 220848786, 1518435648] -> 260282679
[1821793358, 3054510267, 1906284293, 3380915352, 2593544407, 3915598365, 216223976, 4179796969, 2252681786, 35956624, 2373486206, 877849443, 3555235079, 3667331109, 2387597923, 2892840935, 1697182374, 4016309999, 3498537950, 2841634562, 3522184350, 1705717875, 4162597164, 2056807023, 2211204973, 3267303967, 3919558114] -> 325889376
[1201670212, 746140380, 1320204973, 751872422, 3310421711, 1963946901, 625374447, 2435010607, 3055683554, 3426449636, 4008426547, 337682689, 4056318014, 2315028303, 3896431367, 3153777187, 595649858, 3945299365, 113301178, 437820390, 2219035549, 284287245, 3156511236, 3954981533, 599197497, 3946167954, 3728124202, 733618224, 2033404503, 2973733622, 2705618778, 3966553008, 4036118888] -> 389267947
[1910148876, 4015517082, 1214608849, 988091919, 2525094197, 81939600, 3978887728, 2847303411, 66064807, 3605598571, 819502713, 233898591, 1970756897, 2621444751, 213250142, 3439913424, 1332746382, 4063692407, 869169264, 660008864, 2526902458, 3155477426, 1686496804, 3877857092, 4248058769, 2220603751, 465407251, 3418337662, 362811108] -> 424148743
[2663034465, 347641133, 1266329350, 494434286, 24416336, 1697535896, 233051569, 1134030615, 821790629, 1645020605, 2310228737, 63596966, 3832219398, 465243130, 4070794284, 1062758865, 2070536142, 2272514237, 440192702, 3490367376, 3000125608, 1175130963, 339129292, 1200804986, 3560746354, 1043913893, 489469329, 3588157041, 415367051, 1506938118, 519839008, 3921307375, 1017013313] -> 481818733
[2125426152, 580419999, 1942656966, 733532664, 1002041383, 4276401064, 3164003602, 1600445806, 87486541, 585241213, 1822489492, 3670405099, 2371866217, 293071217, 1698095716, 2860716495, 4096732631, 860271933, 1698519068, 1453446230, 2296564613, 1034887983, 3124580892, 3613882243, 3142206150, 1201496338, 262241539, 2658580629, 75299337, 1543495990, 1079614000, 1988996951, 1393947112, 3686457188] -> 503218192
[849928291, 3109119200, 641398141, 953565598, 1651963094, 2973763708, 2409024299, 2638872282, 3438675501, 2024561790, 2574711112, 2224233243, 2145875524, 2483785498, 832036542, 2756591965, 804169437, 1610259885, 1727659540, 3804671151, 2651460465, 3679117113, 3392652401, 4291100134, 3315401681, 4150630548, 3255282492, 1793428708, 4213220288, 1275054855, 3395683809, 126422719, 747072718] -> 529806420
[3795382423, 2318193600, 2941259533, 1429473204, 3841712189, 908410440, 2163447895, 2266401416, 1962272288, 2172290597, 903602965, 2493475709, 1557270830, 1269693466, 1882416966, 1675172531, 2727910931, 67882201, 982110502, 922191612, 1147478737, 1527639226, 1162436584, 3427111171, 1012039523, 775426178, 2235111095, 838473449, 3547292154, 2331447259, 2482093779, 3522096001, 978885775, 911597256, 4199754157, 884328339, 3958245046, 3701693539, 2205845644, 737996384] -> 594670421
[2281257517, 3184614238, 760843542, 2349085191, 3946149756, 3843515781, 357391440, 280771960, 1523563345, 1988608601, 3428233753, 100999796, 1858706576, 3053822297, 2013347998, 2869736987, 3049191880, 2683304507, 2165662057, 1644801829, 3174344328, 516799881, 2817748912, 680940874, 60488796, 440310808, 4238771580, 3333869324, 3437732706, 2291819087, 2904810286, 643100619, 2444662123, 2814934785, 3198564042, 3050641347, 2644737611, 1044358220, 2729428974, 2696207835, 1949284287, 2602160488, 4077269219, 843097934] -> 608761482
[2459044101, 3628291690, 21901316, 1607776779, 574126446, 2632595613, 1898457307, 1993218217, 2268071755, 2262630078, 1334308000, 2756814507, 1322265974, 493521136, 3513003023, 1998755513, 501729387, 461623974, 2703701229, 3074392802, 2895338933, 1257991758, 1147599903, 2188879650, 2802233695, 1049885503, 67969151, 3855902849, 1481701505, 2592817148, 3215907335, 3924969407, 91996333, 857125517, 3408363147, 4006870634, 998389513, 1316925910, 1621658875, 2583060764, 3809067813, 4122609293, 3970592027, 529557207, 1717219729, 2504378950, 418151974, 1944676036, 4136632810, 2747386509] -> 672575322
[58833619, 2580439730, 3693390914, 2286941505, 2380251005, 860188183, 1775128121, 1286129564, 1885665593, 2208299577, 4157266729, 3426550994, 3918528183, 4277973044, 3399609402, 1027156209, 2127983039, 4014806283, 1680366681, 2338809246, 2450101289, 2369487851, 2470761981, 4146252501] -> 758102662
[1228778451, 3161629785, 956090791, 2661866534, 3797423759, 387813121, 686718911, 1671912681, 3122216050, 3708361254, 2450007821, 562470295, 2371567740, 594715145, 1539207036, 1044745741, 3799733486, 3957279143, 3072889407, 43156136, 564713293, 3478075046, 3706967146, 3687044887, 992776898, 3807405066] -> 829276193
[4181681092, 472808862, 2820379353, 1971138909, 957029949, 383690688, 1619203644, 40060279, 799380031, 2522476881, 3052161194, 1101783645, 550394884, 1003243703, 1139176364, 637909704, 1746995412, 3569010375, 2962331845, 1221693446, 1622613434, 2503259347, 1956882342, 3173044265, 1239439145, 842693462, 2538205226, 2451274505, 4212820092, 3591867315, 1233020299, 978021001, 3148810648, 1811827712, 2638973530, 815364991, 1169900461, 1109925685, 856009157, 3254416301, 2037813477, 3945026418, 1535198958, 4259784019, 573350683, 1682977388, 1817703645, 2576589136] -> 898248134
[3182465441, 38281954, 424174823, 2654611156, 2877159272, 723093605, 1443837986, 3392781854, 2463922103, 4286449851, 538962217, 2130457280, 2178815165, 3486677232, 4058242555, 491907811, 162839736, 1759357975, 870141283, 1495756753, 1216746993, 1937724479, 1514235325, 4184006461, 3436845854, 2401877503, 610797147] -> 913459250
[1149034119, 2642531102, 1662163608, 3290983280, 819699523, 515983953, 2696922239, 2166479741, 1179733281, 2022256392, 1626594590, 543353369, 3692263091, 2392093428, 4014491227, 1041518429, 1713650051, 3538550903, 3376348599, 4277108217, 2867735513, 82177527, 10764183, 2126826329, 1028783725, 2271873691, 213268477, 2609718986, 3495904404, 913207638, 1265558725, 2980124058, 303469040, 253105205, 495410356] -> 931867201
[1077541173, 465200833, 3196368211, 1288748206, 2815478016, 1065298665, 2884065004, 392702413, 3587996620, 2894220485, 3226548573, 145951050, 956741233, 2456765795, 3978933657, 1007424556, 2169150234, 3651822773, 1565891297, 510586439, 2219790712, 422126653, 1640535295, 2992880885, 3691821384, 1555208492, 2642710071, 2625348100, 578265950, 592730275, 2092195932, 2504727970, 2149376041, 3179917702, 141987426, 2680075044, 656985656, 2153196909, 4149347932, 2216899428, 2639954700, 3190665042, 2786390314, 2974431514, 1505784370, 3602363995, 1699665616, 327189341, 335160466] -> 945442081
[1284631810, 3717268552, 2270237312, 1334241593, 1477854231, 3138477902, 3837857694, 3586498067, 1959233909, 3891506078, 2756512687, 2553842301, 964459746, 1395510385, 1185295577, 3096625618, 3145085801, 2953203657, 412902939, 3850708910, 2893328412, 2046860068, 86811187, 1188173782, 1294786028, 3044338056, 1040690464, 2128428491, 2571424043, 467837042, 3752155615, 4020048256, 3119445970, 2240289832, 1491368301, 3651270856, 2456196179, 378459336, 188104463, 777143930, 1824526143, 4060407078, 3866701738] -> 1005490735
[329511653, 2198540059, 2154958918, 378488687, 1460553211, 2529097663, 2295211009, 68583162, 1430570391, 1601101442, 1075009348, 2795773614, 833830658, 903181786, 633114581, 3292534356, 3245176784, 2582500259, 3943169203, 1130082435, 3079387854, 3150805871, 1721482617, 2291158702, 3225790458, 1468813727, 862011357, 4268175908, 3661137987, 464141661, 3956758980, 353701536, 1341361227, 809582799, 3968372465, 4111098095, 2898048807, 1509470117, 3255194829, 1330296075, 3520527115, 3736908846, 168999059] -> 1056211089
[1981947965, 3374825360, 3953254353, 1846368367, 1512467473, 3858309918, 4274851313, 787575622, 1876080592, 73919936, 3951066372, 730226758, 3289279637, 768907525, 1773361296, 821210337, 823248190, 869737395, 1360958003, 37854528, 426098214, 2732039516, 64754302, 3276807872, 3816769165, 991587659, 2262236344, 2500333010, 544193426, 1576372318, 2228843007, 2361635236, 2149913894, 3651360213, 2399010654, 4269441712, 1500948518, 3962883634, 1885037482, 2608182179, 3516218188, 513897987, 4242749462, 3892663321, 3922003874, 2123606394, 1383207752, 2867973529, 3840556114, 2969243741] -> 1071722183
[2514263486, 742731713, 3234620175, 3101660088, 2571087369, 3985627296, 3534135449, 2957356517, 2720311929, 2985074784, 2621199558, 3407431088, 2711831197, 1247421057, 3794320157, 1180925811, 1569942151, 2629267152, 173488457, 1406704148, 3969219352, 291552269, 1333806964, 2203738485, 1554560925, 496218017, 1379546509, 2539796280, 4118468566, 2328295795, 884528669, 123600717, 1597383076] -> 1093875318
[1521547521, 982639096, 3108583938, 3860926389, 1477036165, 2087376039, 953332718, 1432039007, 836148138, 3081500185, 396473811, 1621365414, 3796659618, 3939606659, 4248871833, 2917576264, 2729341943, 2981171481, 2311970081, 150264587, 2215220955, 995666639, 452785817, 758560294, 2134005122, 39617150, 1283053470, 4198921194, 2492185018, 595457083, 1042642207, 545670342, 4068770955, 749869737, 984770127, 663196981, 1849907933, 1443537561, 3860175505, 1016163998, 3260982442, 3769639327, 2728953690, 1556597597, 1877966580] -> 1155984918
[1198705327, 1935517658, 503589232, 1577874702, 627636752, 3338218922, 1157893341, 3491377096, 158028724, 2472091015, 1488124620, 2401465711, 3064309912, 1501000204, 3549612642, 2584717560, 1744885213, 4114571438, 1885581118, 2912594621, 3206412254, 4218195409, 2855464167, 2517461088, 685010709, 2000660923, 3562767265, 1193468772, 1470300123, 819551295, 2007863203, 980161306, 717575858, 2022655390, 1011703442, 1190149434, 182438707, 4158211886, 2175561856, 3060913999, 1239412823, 2397907962, 3942897187, 2435381198, 4007365042, 1591831562, 1664232451] -> 1198085477
[3487615865, 4016084043, 1447817513, 2774535999, 1987172957, 3906060658, 3082008686, 88752056, 1947416015, 3376238098, 3503927625, 342228658, 2709452172, 1070052314, 3791872618, 1290870698, 587900606, 1133043554, 671196454, 3876594873, 526665632, 165876551, 3122917378, 4053700531, 1298948490, 1351144611, 563691115, 322287271, 2461185743, 3663527460, 4222266716, 3553092834, 4168116886, 1291668088, 2033147092, 17653398, 3675413239, 2758765766, 2242336270, 734645239] -> 1234710497
[4199109099, 1728109096, 1057011054, 312030739, 490158329, 2571270959, 1337574406, 3037342898, 2943362559, 468588625, 1215641907, 891368463, 3446982561, 146537280, 1406572684, 1996341215, 1239450385, 3239123491, 2654855274, 584040319, 664433665, 3282777560, 3498456462, 398718226, 3318895203, 4109932784, 435092929, 3761442422] -> 1281039914
[330441100, 3682376585, 1548511518, 1012361554, 1641750013, 1347050521, 2746331107, 363031882, 3664742187, 263887825, 3045968855, 4047861721, 4201376344, 2013802400, 43925321, 2057073856, 974221675, 1240944962, 3823005386, 3154676709, 1649841356, 3211264096, 355417272, 2640980835, 2706339938, 987877841, 110633642, 4200784138, 3756125284, 471091914, 3370617977] -> 1307906352
[3242988428, 4255969082, 867271806, 415292052, 3609803438, 416630352, 3232953772, 4269951801, 350204559, 4181264343, 2332795572, 2668659134, 2628956236, 3596517643, 2454368561, 1369266677, 3672701863, 388613580, 1479812495, 2141142524, 2486572053, 2320908196, 393062185, 1938229604, 231136143, 1626064567, 1493467817, 1983339349, 894785483, 44315897, 1514887732, 1864793575, 3953707615, 3492718672, 3522516821, 289358385, 2227856486, 256540739, 4213343606, 200728817, 3396994332, 3845507785, 220892851] -> 1358830177
[3696325595, 95369954, 135918350, 487636311, 825705453, 3049392697, 841126004, 1979734631, 507462131, 1964376824, 3031319264, 3487421046, 2192402845, 3792297163, 3808487536, 1516405605, 664293073, 613658148, 4063050354, 1921745034, 2258497732, 3473698642, 3256102715, 1916456581, 1185919167, 3315712023, 2955859300, 1784025278, 2641317925] -> 1394873382
[2775177776, 3943074566, 1008831348, 423577731, 2936039382, 3419907541, 3346657170, 1767888023, 2549910502, 1510292917, 609091982, 3619041655, 1071140841, 4210891014, 4197129534, 4002247218, 1785145415, 773140313, 850187206, 3695148850, 1699963321, 2029501148, 2273345989, 2045294772, 2490424114, 3670522622, 1903814852, 2845243988, 547217989, 814736327, 1178316093, 3421797242, 583834194, 560861293, 1728122240, 1143281622] -> 1418337751
[3909221229, 1543378013, 3578599386, 1821301593, 382025872, 3090149061, 2799832135, 2969167740, 3214053618, 2459088339, 61851869, 2988154510, 3405155263, 1143312722, 3055626916, 1306722473, 2424970867, 3838250998, 2541260784, 3634562987, 91009990, 228167376, 144739158, 69758376, 1416179539, 175832649, 1058618668, 3141319060, 2603813691, 3448922532, 3563483339, 2187253777, 3939459963, 729053170, 348198400, 2818894106, 4163690148, 4269596985, 1550179869, 2888474224, 977138852, 2267335031, 1039633247, 331897176, 3493431132] -> 1440945260
[548460784, 2008846553, 4095226690, 2537987040, 1412322911, 295388541, 388770296, 2073124976, 4254588769, 2032785043, 3714052855, 653890260, 3378524009, 4169809326, 1850329785, 2009334592, 2701527530, 260935905, 1081271463, 3349408130, 187425716, 588199859, 1037983019, 3586620795, 4254384458, 3345512384, 1424865889, 1577623131, 3897199153, 2727265103, 3204931862, 4089210973, 1909632078, 3588030047, 1946727191, 502364265, 2290334448, 2873594561, 1542446006, 3459132329, 4292608386, 3741690622, 165647068, 1253913368, 1993997790, 2812680991, 2352267270] -> 1500826128
[888893887, 2614777345, 2033163377, 4036989271, 3569896418, 1268962278, 1146226964, 107044610, 2147272927, 3436615686, 1202308143, 1325516848, 2994121867, 973196789, 3793281340, 3099942299, 2472036270, 1798140896, 2858254867, 2204354371, 3472459738, 826142039, 4066377434, 2519409484, 1777413266, 3493671143, 551412865, 886786197, 3882384707, 1172607108, 636563066, 190968362, 3994307050, 3757375918, 2755608252] -> 1566530912
[2820693508, 1716694240, 2268887698, 1296436510, 2916995305, 3041798596, 3096982293, 177546898, 3065170162, 933607211, 102294817, 1801218436, 102569645, 2901869379, 2714886480, 119387632, 656198271, 1388322584, 2775933520, 255861774, 653410879, 121504271, 3607745288, 2230556561, 2594802791, 1189055855, 1111240008, 2530579856, 936965766, 3263165681, 1646076782, 3215192313, 2313499990, 1836078863, 76222274, 3537625615, 3448299995, 1142686967, 2783169043, 1481999457, 407759227] -> 1596242054
[2829936706, 3129475899, 3747369794, 3436426639, 2221019177, 2711897885, 2281503984, 3142203725, 1105679393, 2240453588, 21256724, 1976207648, 2233823367, 3867572431, 883054004, 3828587385, 55294132, 3869553795, 815613304, 1561650320, 1751491302, 3021430621, 3344018026, 1092797427, 2270330908, 872186202, 1096541534, 2148792176, 3302750004, 1221765447, 2805795772, 779878772, 758662454, 4154291008, 1400703216, 2661407119, 3653090591, 674866675, 688479370, 757212805, 2546987102, 3742892714] -> 1598478846
[1350161655, 3410944815, 579310510, 2035583578, 2536805685, 2120035014, 3911683567, 137430281, 676637211, 80424819, 2031627279, 1805522976, 2901778769, 2558787384, 2175054, 1588147737, 1750628274, 1642903038, 2323086809, 204561611, 3017703970, 66172789, 3753895397, 1082808709, 343477515, 3245691790, 2098802152, 2796851615, 3416307802, 1023857834, 1965832789, 412854529, 354175981, 3277463158, 2311727857, 3748028931, 75711999, 605941926, 4268457500, 1591192393, 2493456550, 1285360546, 160198834, 3576865205, 188890618, 1853564214, 4281264410] -> 1682874144
[1759970266, 4134439683, 1532925684, 269299343, 4192017837, 3267142052, 2227585380, 917508171, 2947539107, 1298614074, 3027970270, 3512964090, 78445334, 3939961135, 1550341966, 891964846, 882505655, 314334452, 536698283, 4073297092, 1362911951, 10583051, 4102282272, 39722618, 817646395, 635913246, 3395599638, 114835278, 2721016640, 1301397743, 3359042468, 1967603896, 1842991024, 797964890, 4075738194, 1833629245, 222121931, 3323140076, 3669797521, 2149358182, 3096189598, 4128620833, 1053919363, 3716968416, 1538927892, 3456625069, 3320651259, 2424645618, 3487090498, 1836457156] -> 1757912108
[3206311438, 3995119122, 875115076, 1697246932, 1819127701, 3119338948, 4287411440, 398614242, 2014590172, 1247230330, 4290547991, 2948954546, 3744789096, 2104418209, 185774855, 4253667876, 2202426919, 3397365003, 2934331500, 2095666656, 996668853, 249650538, 3668868008, 1615410768, 492066747, 3451532273, 3886898259, 1151021320, 1782145434, 2431016438, 2907780443, 2152384331, 800661361, 2028554209, 1103300513] -> 1779025760
[1744880977, 2233651387, 56617556, 1115479166, 2611598968, 1306003390, 2910946867, 1496609711, 4077408131, 3512362374, 2281393975, 3418144659, 1773074316, 1601343734, 807592751, 2365582522, 1446242874, 847706414, 2850851126, 414921605, 2617956981, 4205647553, 4063233713, 142946783, 1975643220, 1993142296, 3392343515, 1515671667, 1948071803, 3991728328, 968499177, 2198785340, 1321024334, 3781705466, 3920003634, 727611412, 30865083, 1239101335, 3511359132, 3435471655, 2332994209, 3451243757, 2388153527, 3219135587, 3239646757, 130935018] -> 1815085532
[3742736007, 3013982958, 950712805, 590882231, 2021027712, 1687735825, 2840162788, 236595715, 1966708028, 3189262610, 1430349774, 533579944, 2649611396, 2583295302, 3582250057, 2565346361, 396439695, 4153603028, 3342863641, 209226977, 3673492099, 1470755570, 1265241721, 755285979, 1871568539, 1835865540, 614068679, 1981443993, 570207739, 4057872936, 636526617, 3083893122] -> 1892006865
[2363102448, 2664740190, 3782032362, 2321087205, 3789479744, 439476800, 267092854, 559662253, 3052920007, 1039904017, 882951865, 1185772050, 121779049, 1255320891, 3898061704, 4160716348, 1870139159, 1739210570, 2851269822, 866317206, 559870442, 3322246740, 419503690, 1438071640] -> 1972485478
[3678708758, 283072981, 1128708041, 648928081, 516617525, 2562088844, 2077772312, 454404108, 2773100394, 3731939823, 3956390266, 1273038228, 711119494, 1889950256, 2150587570, 2368988611, 3083613740, 136402153, 3706807002, 1951409673, 646402003, 1965112544, 3799779787, 2727912100, 3934407652, 3190480389, 2893737274, 3503914114, 1242998415, 3673986521, 3585553888, 1356707071, 2892703448, 2503698089, 677538230, 3777556915, 4131625633, 3989648217, 3371399362, 916533795, 438016809, 1888923541, 4018783570, 35576878, 1439895305] -> 2006012423
[4186976930, 1468422130, 2127537262, 3013431126, 1014175853, 619237775, 2251537150, 2635437352, 2529830434, 270079805, 2249885826, 4160741826, 1799997550, 90064691, 1707888518, 4191506978, 1400211336, 4260528976, 1777462871, 3522395605, 3184921911, 1289756682, 1204481339, 2320228865, 1382458891, 2585094069, 1807393027, 904952345, 2418050161, 3944759600, 850396791, 2010841661, 911793010, 2814160212, 3917390566, 1742258451, 312330630, 1509247515, 3622596041, 2381664940, 792997581, 1348701222, 4231392479, 1492858603, 841395446, 1732373831, 2792829070, 1232839054, 4133086979, 1330710519] -> 2023628655
[1447486536, 1589601829, 2711407664, 2671174482, 755815460, 2739822893, 624189525, 3881531582, 2104935742, 3877151275, 4029158122, 3270473853, 2397135808, 541842801, 4159848352, 1265295340, 2924665155, 1697146316, 1323336501, 1887970585, 3957105141, 4080944661, 1260974434, 2538893190, 1418098060, 1087907449, 1555085135, 517760467, 1165667983, 1976195378, 899913113, 3298189181, 2844731428, 2587250768, 419676280, 3669952139, 1008835697, 2109855126, 2204851528] -> 2080799654
[3057502862, 2753037643, 586428454, 875470641, 763920856, 1915854920, 960628797, 3268824212, 155894276, 326435370, 4153187586, 3266725420, 3421441761, 984314678, 134133671, 1571569456, 3290768531, 1270814441, 3360280927, 63041059, 1367587866, 3837387650, 3016828154, 1932299196, 2677431226, 118282642, 4274320011, 2842715068, 3590275349, 3385213492, 3669399717, 23889551, 1990874337, 2065027029] -> 2124973901
[404963327, 970175530, 2315604326, 759401990, 3265128006, 3886300253, 2428903513, 2547895316, 2394147688, 1256335272, 54823458, 1121206884, 1587348808, 1813532729, 475467047, 2240315018, 4228484438, 3836003837, 1713780997, 1021495744, 1667817101, 3431813061, 1741019894, 244489142, 4185223756, 2154392407, 1130398867, 2200660986, 1655353090, 824162463, 269754245, 853115597, 554916755, 3901389184, 3490586038, 2387525212, 3363338561, 2848158299, 1365520662, 65222054, 969608783, 3832404018, 797302386, 220365549, 998915816, 3880644358, 1656707899, 113713704] -> 2200030708
[1182658594, 858230724, 1872357797, 2818651205, 2361035630, 2882362708, 774158555, 1780906106, 1245895866, 1925774739, 1695790856, 1650553385, 3918997115, 1289270838, 1080124671, 1854179057, 4032572199, 3398178415, 2660852575, 2126762197, 2845072225, 3916118969, 2961311910, 3490728772, 2152765282, 925727050, 2046977563, 3481846724, 3016398740, 3170193754, 3947157907, 2318811664, 665856332, 2694492251] -> 2222600201

• Do you have a reference implementation which does this in a reasonable time? I know that I've tried reversing MT with the Z3 constraint solver without success. – Peter Taylor Nov 27 '16 at 14:40
• @PeterTaylor I don't have a solution myself, though granted I haven't put much effort towards coming up with a solution yet. – user45941 Nov 27 '16 at 18:29

# Implement the Terducodec

In this challenge we'll be implementing a tersum duplicate coder/decoder; aka, a terducodec. This is a reciprocal substitution cipher, so the encoder and the decoder are the same.

## Definition

The tersum duplicate of a number x is 2x using base 3 without carries. Alternately phrased, the tersum duplicate of x in base 3 is simply x in base 3 with the 1 and 2 digits swapped.

For example, the tersum duplicate of 47 is 64, because 47=1202b3, and 2101b3 is 64. Note that since this is simply a result of swapping base 3 digits, the tersum duplicate operation is its own inverse.

## Terducoding

We code at an octet level, using values in the range 0 to 255 and the tersum duplicate operation. To keep the result in the range 0 to 255 the operation is applied to each of the following subgroups:

• 0 through 242
• 243 through 251
• 252 through 254
• 255

Given n in the range 0 to 255, let s be the lowest value in the same subgroup as n. Then n can be terducoded as follows:

• Subtract s from n
• Calculate the tersum duplicate of this result
• Add s back to the above

## Examples

The value 195 is in group 0 through 242. This group's s value is 0. Thus, we simply take the tersum duplicate of 195=21020b3, which is 12010b3=138.

The value 246 is in group 243 through 251. This group's s value is 243. Thus, we take the tersum duplicate of 246-243=3=10b3, which is 20b3=6, and add s back in to get 249.

The value 252 is in group 252 to 254. This group's s value is 252. Thus, we take the tersum duplicate of 252-252=0=0b3, which is just 0b3=0, and add s back to get 252.

## The Challenge

Write a program or function that implements the terducodec. Your program or function should accept as input an ordered sequence of octets, and produce as output a corresponding ordered sequence of terducoded values. The input/output mechanism may use standard input and standard output, arrays, etc; however, your chosen mechanism must be able to represent any sequence of octets where each octet has the full possible range.

This is code golf. Standard loopholes disallowed. The smallest terducodec in bytes wins.

## Test Case

A single test case doubles as a list of all possible encodings. Here, the decimal representation is used for each octet.

  0   2   1   6   8   7   3   5   4  18  20  19  24  26  25  21
23  22   9  11  10  15  17  16  12  14  13  54  56  55  60  62
61  57  59  58  72  74  73  78  80  79  75  77  76  63  65  64
69  71  70  66  68  67  27  29  28  33  35  34  30  32  31  45
47  46  51  53  52  48  50  49  36  38  37  42  44  43  39  41
40 162 164 163 168 170 169 165 167 166 180 182 181 186 188 187
183 185 184 171 173 172 177 179 178 174 176 175 216 218 217 222
224 223 219 221 220 234 236 235 240 242 241 237 239 238 225 227
226 231 233 232 228 230 229 189 191 190 195 197 196 192 194 193
207 209 208 213 215 214 210 212 211 198 200 199 204 206 205 201
203 202  81  83  82  87  89  88  84  86  85  99 101 100 105 107
106 102 104 103  90  92  91  96  98  97  93  95  94 135 137 136
141 143 142 138 140 139 153 155 154 159 161 160 156 158 157 144
146 145 150 152 151 147 149 148 108 110 109 114 116 115 111 113
112 126 128 127 132 134 133 129 131 130 117 119 118 123 125 124
120 122 121 243 245 244 249 251 250 246 248 247 252 254 253 255
->
0   1   2   3   4   5   6   7   8   9  10  11  12  13  14  15
16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31
32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47
48  49  50  51  52  53  54  55  56  57  58  59  60  61  62  63
64  65  66  67  68  69  70  71  72  73  74  75  76  77  78  79
80  81  82  83  84  85  86  87  88  89  90  91  92  93  94  95
96  97  98  99 100 101 102 103 104 105 106 107 108 109 110 111
112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143
144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175
176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191
192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207
208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223
224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239
240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255


# Minimize Your Code's SHA-256 Hash code-challenge

Write a program or function that outputs its own SHA-256 hash as a hex digest. The winner will be the program with the smallest output (when interpreted as a hexadecimal integer).

## Rules

• Accessing the program/function's source code is allowed (a la cheating quines).
• The output must be in hexadecimal, optionally padded with zeros to 64 hex digits.

## Example (Python 2)

import hashlib


Output/score:

fe58e530aa530b34535cadc001383750a3343634c644539901a04659b41a7308

• What's the tie-breaker (or real scoring criterion)? Most languages which can do the challenge should be able to score 0 on the current criterion using a 9-year-old attack on MD5. (Alternative: change it to use SHA). – Peter Taylor Nov 29 '16 at 9:02
• @PeterTaylor Fair point. I'll change it to SHA-256. The tie-breaker is earliest submission to achieve the score, as is default. – user45941 Nov 29 '16 at 9:03
• Will this turn into a brute force competition by changing a single letter to find the lowest hash? – Angs Nov 29 '16 at 17:55
• @Angs Though that is one strategy, I don't expect it to be the most fruitful strategy. – user45941 Nov 29 '16 at 19:36
• I'm pretty sure this is just a brute-force competition. I don't see any reason why changing the coding part of the program would help, so you'd just brute-force the content of a comment needed to minimize the hash value. As such, I don't think it's an interesting problem. (Even requiring proper-quine rules wouldn't help; people would just use a string literal instead.) – user62131 Nov 30 '16 at 0:30
• @ais523 Do you think that a maximum byte count (like 256 bytes) and proper quine rules would help? – user45941 Nov 30 '16 at 0:35
• No. 256 bytes is easily enough to add in a fairly long comment or string literal alongside the program, and proper quine rules would simply disqualify languages where proper quines are long, whilst not otherwise affecting things. – user62131 Nov 30 '16 at 0:41
• @ais523 I'm interested in salvaging this idea, so if you have any suggestions for improving it, I'd love to hear them. – user45941 Nov 30 '16 at 0:41
• The only thing I can think of is code-golf with the hash as the tiebreak; code-golf naturally prevents people using unnecessary characters, so it would come down to picking whichever of equivalent versions of the program had the lowest hash. – user62131 Nov 30 '16 at 0:48
• @ais523 That's not a good idea. Besides, it would be a duplicate of the current SHA-256 code golf challenge. – user45941 Nov 30 '16 at 3:37

# Shooting range

Print/output this exact text:

                   _
/ \
/ _ \
/ / \ \
/ / _ \ \
/ / / \ \ \
/ / / _ \ \ \
/ / / / \ \ \ \
/ / / / _ \ \ \ \
/ / / / / \ \ \ \ \
/ / / / / _ \ \ \ \ \
/ / / / / / \ \ \ \ \ \
/ / / / / / _ \ \ \ \ \ \
/ / / / / / / \ \ \ \ \ \ \
/ / / / / / / _ \ \ \ \ \ \ \
/ / / / / / / / \ \ \ \ \ \ \ \
/ / / / / / / / _ \ \ \ \ \ \ \ \
/ / / / / / / / / \ \ \ \ \ \ \ \ \
/ / / / / / / / / _ \ \ \ \ \ \ \ \ \
/ / / / / / / / / / \ \ \ \ \ \ \ \ \ \
| | | | | | | | | |x| | | | | | | | | |
\ \ \ \ \ \ \ \ \ \_/ / / / / / / / / /
\ \ \ \ \ \ \ \ \ 9 / / / / / / / / /
\ \ \ \ \ \ \ \ \_/ / / / / / / / /
\ \ \ \ \ \ \ \ 8 / / / / / / / /
\ \ \ \ \ \ \ \_/ / / / / / / /
\ \ \ \ \ \ \ 7 / / / / / / /
\ \ \ \ \ \ \_/ / / / / / /
\ \ \ \ \ \ 6 / / / / / /
\ \ \ \ \ \_/ / / / / /
\ \ \ \ \ 5 / / / / /
\ \ \ \ \_/ / / / /
\ \ \ \ 4 / / / /
\ \ \ \_/ / / /
\ \ \ 3 / / /
\ \ \_/ / /
\ \ 2 / /
\ \_/ /
\ 1 /
\_/


Trailing whitespace after line ends allowed, trailing/prepending newlines also are. You can write either a full program or just the function. This is , so shortest code (in bytes in your languages preferred encoding) wins.

### sandbox:

• any better wording?
• anything I'm missing?

• this would look better with _ instead of - at the top, especially in normal spaced text (as opposed to the wide line spacing used on SE.) You would need to shift the _ up one line, obviously. I also think maybe a more regular octagon would both look better and be more challenging than the current near-diamond shape. – Level River St Nov 28 '16 at 21:23
• @LevelRiverSt You're right that at the top should be underscores. The reason I had it like that is so the top and bottom part would be mirrorable (sort of) and would increase the complexity but also reduce length of code. I don't really like the idea of an octagon as a big part would just be the same. – dzaima Nov 29 '16 at 14:38

# Paint the Path

First shot ever on something like this...

You got stuck in snow and have to get out somehow. You only have a paper left with some letters on it signing the directions you had to go marked with NESW. To make sure to not get lost you want to make yourself a map with the pathing. You grab your laptop and start coding. After you are done you try your function/program and check if it works as intended. It does! You grab yourself thick clothing and a shovel and work yourself through the snow.

Your task is to write a program or function that takes a series of characters into your standard way of input. This might be a string, a list/array of characters/strings only consisting of the letters NESW and ouputs a map with the characters ^>V< as arrows signing where to go. Each sign has to follow directly after the one before at the correct position (>^ is correct ^> is not).

## Input

You can take the input in a way of your choice as described above.
You may assume that the way will never cross.
You may assume that you will never go further North nor further West than your house is.
You may assume, that you will never have to walk back directly (eg SNWE).
You will always start in the top-left/north-west corner.

## Output

The output should represent a kind of a map with the characters^>V< as movement signs. You may use spaces or dots to separate to quickly find the way.

# Examples

Input: EEEE
Output: >>>>

Input: SENE
Output:

V>
>^


Input: SEESENESSENNNE
Output:

V    >
>>V>V^
>^V^
>^


Input: EEESESWSSSEEENNENWNNEEE
Output:

>>>V  >>>
>V ^
V< ^<
V  >^
V  ^
>>>^


## Scoring

This is so the shortest answer in bytes wins!

## Sandbox questions

Has there already been something like this?
Is the story redundant or acceptable?
Did I miss something? Like a requirement?
What tags should be used (except for code-golf)? I am not familiar with the tags and welcome any suggestion.

• I think the third example is missing the first V – Leo Nov 30 '16 at 15:10
• @Leo Indeed! Thanks for the catch :) – geisterfurz007 Nov 30 '16 at 15:14
• and looking again, your last example seems even more messed up :) – Leo Nov 30 '16 at 16:09
• @Leo Errrrm... Something (most likely me...) must have messed up the formatting... Should be fixxed now and thanks again for your help! – geisterfurz007 Nov 30 '16 at 16:27
• ehm... no, it isn't :D (or at least it doesn't follow the directions) – Leo Nov 30 '16 at 17:43
• @Leo zzZzZzZz Forgot to change it, after changing it in the program, I tested with. I should not do something like that after work, I guess. Thanks for each of your correcting looks. – geisterfurz007 Nov 30 '16 at 18:15

# Self-Verifying Quine Variant quinecode-golf

We have plenty of challenges that ask you to output your program's source code, but I couldn't find any where you input your program's source code and verify it.

## The challenge:

Your code is to accept some text as input, and return truthy if the input is identical to your source code, and falsy if it is not identical to your source code.

For the purposes of this challenge, your code need only verify against the actual code that you typed, not any runtime flags you used or implicit code the compiler/interpreter adds on.

Standard loopholes are disallowed, and standard quine rules apply (no reading your source code from disk).

This is , so fewest bytes wins.

• It's a good idea, but unfortunately it's been done before. – James Nov 30 '16 at 23:12
• Ah well. It was good while it lasted. – Gabriel Benamy Nov 30 '16 at 23:13

# Wordsearch programming

(Todo: example submission + format)

• Write a full program that takes no input and prints any message under 100 characters to STDOUT.
• This program must print the same message every time it is run.
• Arrange your program into a wordsearch puzzle.
• Post the wordsearch and language, along with the desired output.
• Robbers will try to reconstruct the original program.
• If they create any program that produces the desired output from the wordsearch within a week, your submission is cracked and the robber gets some points.
• If your wordsearch survives the week, you get some points.
• You must reveal what your original code was after the week for it to earn points.
• Any code snippets in the wordsearch must be at least 4 characters long.
• Whitespace doesn't count in the wordsearch puzzle (int var would become intvar in the wordsearch. The robber can choose to interpret that as int var, intvar, or even in tvar).
• Code snippets can overlap in the wordsearch as long as no code snippet is located completely inside another snippet.
• The wordsearch is case-sensitive.
• There is a cap of 25 rows or columns of the wordsearch.
• No standard loopholes.

## Scoring

Safe submissions earn the cop

100/(number of cells ^ (2/3)) points


For example, a 20x20 submission would receive

100/(400^(2/3)) = 100/54.3 = 1.84 points


100/(150^(2/3)) = 100/28.2 = 3.54 points


Cracked submissions earn the robber

(number of cells ^ (2/3))/10 points


For example, a 20x20 submission would receive

(400 ^ (2/3))/10 = 54.3/10 = 5.43 points


(150 ^ (2/3))/10 = 28.2/10 = 2.82 points


Your total points are the points that you receive from all of your safe submissions plus points from cracking other submissions. The user with the most points by (date a few weeks from challenge posted) wins.

## Example submission

For the example submission, let's say that the program is a simple "Hello, World!" program in java:

public class ictc{
public static void main(String[] args){
System.out.println("Hello World!");
}
}


The first step to the submission would be to strip all whitespace:

publicclassA{publicstaticvoidmain(String[]args){System.out.println("HelloWorld!");}}


The submission would look something like this:

# Java, 10x10 grid

publicvgra
lut.ptotto
n[b;main(S
(]Slatd{}y
"a{{iiints
Hrintcount
egnirts;Se
lssalcrram
l);}}{;;;.
o{World!"o


Output:

Hello World!

• A few notes: I tweaked the scoring quite a bit to get point values that I think can be directly compared between cops and robbers and that incentivize posting smaller puzzles (but not too much). However, feedback about any part of the challenge is appreciated. – Daniel M. Dec 2 '16 at 4:21
• A sample submission would be helpful – user41805 Dec 2 '16 at 17:21
• @KritixiLithos I added a sample – Daniel M. Dec 2 '16 at 22:45
• It's not mentioned, but I assume every character in the grid must be "used" (by both the cop and robber)? Are comments allowed/banned? – Geobits Dec 6 '16 at 16:04
• @Geobits Not every character must be used, but every character can be used. Any space in the word search that cop's program doesn't fill must be filled with a letter of the cop's choice. It is possible for one of these "junk" letters to become part of the robber's program, so the cop needs to choose the filler letters wisely – Daniel M. Dec 6 '16 at 16:21

# Challenge

Your code, which can be a standalone program or a function, should take an input string. You can assume the string only consists of ASCII characters.

The program should then print or return 0 if the text contains more 0-bits, 1 if it contains more 1-bits or something else (see below) if the number of 0-bits and 1-bits is equal.

# Details

The number of bits in a text is determined on a per-character basis: you take the ASCII code of the character (which is 0-127 in pure ASCII) and convert it to 7 bits of binary, then count the number of bits that are either set (1) or cleared (0). The occurrences of these bits are then summed over every character of the input string.

If the number of 0-bits and 1-bits is equal, the program should return or print something else than 0 or 1. This can be returning or printing a negative number, printing nothing or returning null (or similar), etc.

The input will only have a single line and will not have any trailing newlines. You will only need to support the set of printable ASCII characters (ASCII codes 32-126) in the input. If the input is empty, act as if it had an equal number of 0-bits and 1-bits.

# Example

Code Golf


Convert to ASCII codes:

67 111 100 101 32 71 111 108 102


Convert the numbers to 7-bit binary:

1000011 1101111 1100100 1100101 0100000 1000111 1101111 1101100 1100110


The number of 0-bits is 28 and the number of 1-bits is 35, so 1-bit is the dominant bit. Therefore, the program should print or return 1.

# Test Cases

Assuming equal number of 0-bits and 1-bits causes nothing to be printed:

Input: Code Golf
Output: 1

Input: DETERMINING THE DOMINANT BIT
Output: 0

Input: Equal Number of BITS?!
Output:


Remember, this is , so the valid code with the fewest bytes wins. Standard loophole rules apply.

• I'd probably change "should not" to "will not" in "The input should not have any trailing newlines.", and maybe add a test case with non-printable characters that has at least one newline in it if that needs to be supported, since that changes the challenge a lot for languages which by default read one line at a time (e.g. Python). Ditto for \r\n. – Sp3000 Dec 2 '16 at 21:48
• I'd probably say one line of input is enough. Will edit. – dark_blue Dec 2 '16 at 22:15
• Just to confirm, do code points 0-31 and 127 (with the exception of line feed, etc.) have to be supported? – Sp3000 Dec 2 '16 at 22:28
• It's a tough call - I can see requiring support for those cutting off some platforms and languages, so perhaps only printable ASCII would need to be supported. – dark_blue Dec 2 '16 at 22:30
• Personally I don't mind either way, as long as you're clear about it (but yes it would make it harder for some platforms/languages). But if you're indeed going with printable ASCII, I think it'd be better to put the clarifications in the Details section rather than at the end of the post. – Sp3000 Dec 2 '16 at 22:36
• Very closely related: codegolf.stackexchange.com/q/4434/194 codegolf.stackexchange.com/q/6239/194 – Peter Taylor Dec 2 '16 at 22:37
• The Extra section currently only contains clarifications about I/O anyway, so it'd probably be good to just rename it and move it upwards above the test cases and the example. – dark_blue Dec 2 '16 at 22:38

# Piet (Mondrian)'s Puzzle code-challengetilingoptimization

The Mondrian Puzzle (for an integer n) is the following:

Fit non-congruent rectangles into a n*n square grid. What is the smallest difference possible between the largest and the smallest rectangle?

For 6, the optimal difference for M(6) is 5, and can be demonstrated like so:

 ___________
| |S|_______|
| | |   L   |
| |_|_______|
| |     |   |
| |_____|___|
|_|_________| (fig. I)


The largest rectangle (L) has an area of 2 * 4 = 8, and the smallest rectangle (S) has an area of 1 * 3 = 3. Therefore, the difference is 8 - 3 = 5.

Keep in mind that currently, no optimal solutions for n > 24 have been found.

Your task is to create a program that generates a Mondrian grid that contains a (non-optimal) solution, given an integer n.

You will be tested on the numbers from 100 to 150. Your score for each test will be the difference between the largest and smallest rectangle. Your total score is the sum of your scores for all the tests from 100 to 150.

You must present your output like so:

{number}
{grid}


Where number is the score (the difference between largest and smallest), and grid is either:

• A multi-lined string, or
• A two-dimensional list.

The grid MUST clearly show where a rectangle starts and ends.

## Rules:

• Your program must generate the same solution for an integer n every time the program is run.
• You must provide a link to the outputs of all 50 solutions (using Pastebin, Github Gist... anything, really).
• You must have at least two rectangles on the square grid for your solution.

## Meta:

• Is this a dupe?
• What tags?
• Can I improve on my explanation?
• 1. Rectangles, not squares. 2. It would be good to state explicitly that the optimal values are currently unknown for n > 24, and otherwise to use the word optimal much less, because it's misleading. No-one is likely to find solutions anywhere near optimal for n=100 in 10 minutes. 3. The hard time limit, the deterministic output requirement, and the impossibility of searching the space in that time limit are a really bad combination. Depending on system parameters, load etc. the amount of searching that can be done in 10 minutes will vary by an order of magnitude. – Peter Taylor Nov 30 '16 at 8:42
• @PeterTaylor Will it make a difference if I increase the hard time limit? The output is essentially as flexible as possible, and I want a way to make sure that the bots aren't BS-ing about the output (so, something that outputs 1 all the time will be caught under the "output" rule). – clismique Nov 30 '16 at 8:46
• The thing is, normally with a search with time limit what you do is to search, and when you detect that you're almost at the time limit you print your best result so far and quit. As a workaround if I can test on the machine with the time limit I could do a pre-submission run where I find the best scores available in 95% of time and then hard-code an early exit when I reach those scores. But if I'm testing on my 7-year-old desktop (really time to upgrade!) then I don't know how far I could get on your test machine. – Peter Taylor Nov 30 '16 at 8:52
• The normal workaround is to make the final scoring for questions like this be evaluated by the challenge setter on their computer. It's not ideal for a couple of reasons, but at least it means that people can push up to the time limit rather than having to play it safe and only use 50% of it. – Peter Taylor Nov 30 '16 at 8:54
• @PeterTaylor The thing is, I want to make brute-forcing as difficult as possible to discourage it, and encourage different algorithms to generate the grids and answers instead (because that isn't very interesting, TBH). Is there any way around this other than setting a hard time limit? Also, if the time limit is an hour for each number, then it takes 50 hours to get all of the numbers... how would you get around that? – clismique Nov 30 '16 at 8:54
• @PeterTaylor But that idea involves a lot of me, though - I want the challenge to be solved with minimal oversight (because I'm very lazy). It's worked so far for my other metagolf challenge, so I think it should work fine here. – clismique Nov 30 '16 at 8:55
• An answer which doesn't do some kind of search isn't going to be competitive. The competition is going to be in finding a heuristic to guide the search. A time limit is about the only externally measureable limitation. (A memory limit wouldn't work at all). – Peter Taylor Nov 30 '16 at 9:14
• @PeterTaylor Ah, I understand. Is the "same answer" condition alright, in that case? – clismique Nov 30 '16 at 9:16
• 1. Unless I'm overlooking it, you need to disallow the trivial solution of using a single NxN square. 2. You might want to link to oeis.org/A276523. 3. Tags: this isn't metagolf. (Metagolf is generating short programs.) Make it code-challenge, maybe add optimization and probably add set-partitions and/or tiling. – Martin Ender Nov 30 '16 at 13:13
• For comparison, I posted this challenge and this challenge with no time limit, so that people can just run their code for however long they like on their own computer and post their best result. Since the problem sizes were chosen to exclude the possibility of finding an optimal solution, this still forced people to seek better algorithms rather than use trivial brute force. I suppose it's more like "non-trivial brute force". – trichoplax Dec 2 '16 at 10:52
• (I didn't require deterministic output for either of those.) – trichoplax Dec 2 '16 at 10:53
• I've reworded slightly because "solution" seemed to be being used to mean more than one thing. Hopefully that's clearer, but feel free to revert the changes if they don't suit you. – trichoplax Dec 2 '16 at 11:34
• I notice that "Metagolf" still appears in the title. I don't have a better title suggestion, just pointing it out. – trichoplax Dec 2 '16 at 11:42

# Lasers & Mirrors

ah lasers. they're a modern marvel. From the most common colors of blue, red, and purple to the more exotic cyan, green, and Occasional yellow. One of the most fascinating properties of lasers is their tendency to travel in a straight line, and bounce off mirrors. What fun! Let's simulate that.

consider a square grid of arbitrary size. This grid has four elements, Mirrors, walls, spaces, and a single goal

## Mirrors:

Mirrors comprise only 2 characters: / and \. each of them represents a mirror which is reflective on BOTH sides, and is angled at 45 degrees and -45 degrees respectively. If a laser fits a mirror, it will change direction by 90 degrees, as it would in real life. the direction that the laser goes is dependent of the angle of incidence.

## Walls:

Walls also comprise of 2 characters: | and _. A wall is simple in that it stops a laser, and ends the beam. The Direction of the wall is irrelevant. | and _ are functionally the same! If a laser hits a wall before the goal, then the solution is faulty.

## Goals:

There is one goal per puzzle input to the program. The goal is represented by a * somewhere on the grid. The objective is to get the laser to pass through the goal from any cardinal direction.

## Spaces:

Spaces are the easiest to recognize, as they're any character that's not already mentioned. for the purpose of the puzzle, you can assume that all spaces are . (dots).

# Input:

Your input will be a square grid with grid elements and one goal. The grid will always be square, and will always have a goal. grid elements will be entered in a list that is either 1 row tall (eg: the 3x3 example puzzle would be [".","_",".","|","*","|",".",".","."]) or in a square set of arrays (eg: the 3x3 puzzle would be [".","_","."],["|","*","|"],[".",".","."]).

# Output:

Your output will be a number which represents where the laser should be shot from in order to get to the goal. You can see how outputs are numbered below. If there is more than one solution, you may either output any or all solutions.

# Examples:

 ..9
0._.8
1|*|7
2...6
345

Output:4


This example shows how a grid will be numbered. it is numbered counter-clockwise starting in the top left at 0 such as there are 2 numbers for each row and column.The laser is placed at the position of the number and shot in at a multiple of 90 degrees.

The correct answer is 4 because the goal is enclosed by walls except where the 4 is located, which the laser gets shot in (in this case up). In the picture below, you see the same puzzle, but numbered correctly, along with another test case.

the directions on each example show the direction of travel for the laser for each number

This is a code golf, so the shortest code will win! good luck!

• What if there's more than one possible solution? – Gareth Nov 16 '16 at 20:23
• @Gareth thank you. I added to the post: "If there is more than one solution, you may either output any or all solutions." – tuskiomi Nov 16 '16 at 20:25
• Any other feedback? – tuskiomi Nov 17 '16 at 20:05
• If a laser hits a wall parallel to the direction that it's going, the laser will still stop. I'm not sure if this is meant to say perpendicular or parallel. Also, whichever type of wall stops a given direction of laser, what does the other type of wall do? Can a laser pass straight through in that case? Or is the direction of the wall a red herring? Do all walls behave the same regardless of direction? – trichoplax Nov 18 '16 at 20:40
• the compass directions on each example show the direction of travel for the laser for each number Usually "compass directions" would be used for North, East, South, and West. If you want to you up, right, down, and left, it's probably clearer to just leave out the word "compass" and just say "the directions". – trichoplax Nov 18 '16 at 20:43
• you can assume that all spaces are . (dots) so I don't understand the Xs in the last example. – trichoplax Nov 18 '16 at 20:59

# The evolution of integers

The game of life (GOL) is a famous cellular automaton where cells turn on and off depending on how many neighbors they have. For a quick overview, the rules are:

Every cell is either alive or dead. Each generation:

• Every live cell with fewer than two live neighbors dies

• Every live cell with more than three live neighbors dies

• Every live cell with two or three live neighbors lives to the next generation, and

• And every dead cell with exactly three live neighbors is brought back to life.

You can play around with it here to see what happens with these rules.

The cool thing about Game of Life is that the structures and patterns will rapidly evolve over time creating some cool patterns. Let's take this idea of evolving patterns and use it for evolving integers. Here's what your submission must do:

1. Pick a starting configuration of the GOL. We'll call this your "template". Now you must write a program that has the same visual layout as this template. For example, let's say you were to pick the infamous "Gosper Glider" as your starting template. For reference, it looks like this:

 *
*
***


Now your program can only have non-whitespace characters where the live cells are. So in my fictional programming language "Spam", this is a valid program:

 a
b
123


Running this program must output the number '1', optionally followed by a trailing newline.

2. Evolve your program. To evolve your program, apply one generation of the rules of GOL, where whitespace is a dead cell and non-whitespace is a live cell. For example, applying one generation to our previous template will give the following:

 * *
**
*


Now you must write another program in the same language that follows this new template. This new program should print the number '2', again optionally followed by a trailing newline.

3. Repeat this process, evolving your template and increasing the number that is output for as long as possible.

Your score is the total number of generations your code goes while still producing the correct number. If someone manages to create a program that works to infinity, for example by using a gun or something similar, the winning criteria will be the number of live cells in the starting configuration. Earlier posting date will be used as a tiebreaker.

Any questions?

# Print the missing primes

Write a program or function that, when passed a numerical input x, prints or returns the primes beneath the square root of x1 that are not factors of x.

### Examples

Let f(x) be the function called:

>>> f(5)
[2]

>>> f(20)
[3]

>>> f(60)
[7]

>>> f(100)
[3, 7]

>>> f(10000)
[3, 7, 11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 47, 53, 59, 61, 67, 71, 73, 79, 83, 89, 97]


### Bonus Rules

• You may use any builtins that your language provides.
• Your program must support an x input as high as the upper bound defined by your language.

1 Using the square root as only primes below the square root can actually be involved within the factors of x. Without making this restriction, larger numbers would have a lot of excess numbers.

# Meta

More specification?

# 3-way Polyglot Prime Checker code-golf

Make a program that checks if a number is prime in three distinct languages.

• Two versions of the same language aren't considered distinct languages.
• Standard loopholes apply
• neeed... feedback... – noɥʇʎԀʎzɐɹƆ Dec 10 '16 at 14:33