# 3D ASCII Block Building

## Challenge

Write a program that takes an 11x11 array of integers, and constructs a 3D ASCII block building, where each value in the array represents the height of a column of blocks at the coordinates matching the array position. A negative height is a "floating" column - only the top block is visible.

## Example

                                                        __________________
___            /\__\__\__\__\__\__\
3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,      /\__\          /\/\__\__\__\__\__\__\
2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,     /\/__/         /\/\/__/__/__/__/__/__/
2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,    /\/\__\        /\/\/\__\      /\/\/__/
1, 0, 0, 7,-7,-7,-7,-7, 7, 0, 0,    \/\/\__\      /\/\/\/__/     /\/\/__/
0, 0, 0, 7,-7,-7,-7,-7, 7, 0, 0,     \/\/__/     /\/\/\/\__\    /\/\/__/
0, 0, 0, 6, 0, 0, 0, 0, 0, 0, 0,      \/\__\    /\/\/\/\/__/   /\/\/__/
0, 0, 0, 5, 0, 0, 0, 0, 0, 0, 0,       \/__/    \/\/\/\/\__\_  \/\/__/
1, 0, 0, 4, 3, 2, 1, 0, 0, 0, 1,                 \/\/\/\/__/_\_ \/__/
1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1,            ___   \/\/\/__/__/_\_         ___
1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1,           /\__\   \/\/__/__/__/_\       /\__\
1, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1,           \/\__\   \/__/__/__/__/       \/\__\
\/\__\_________         ______\/\__\
\/\__\__\__\__\       /\__\__\__\__\
\/__/__/__/__/       \/__/__/__/__/


## Input

The input will be a list of 121 integers, either read from stdin (choice of separator is up to you), or passed in as an array (can be 1D or 2D).

Heights will be in the range -11 to 11.

## Output

The generated building can be written to stdout, displayed directly on the screen, or returned as a newline-separated string.

Leading and trailing whitespace is allowed.

## Building Rules

The shape of an individual 3D block looks like this:

 ___
/\__\
\/__/


And a 2x2x2 cube of blocks looks like this:

  ______
/\__\__\
/\/\__\__\
\/\/__/__/
\/__/__/


When blocks overlap, a higher block takes precedence over a lower one, blocks in front take precedence over those further back, and blocks to the left takes precedence over those to the right. The only special case is that the top line of a block should never overwrite any non-space character behind it.

The interpretation of column heights can best be explained by looking at a 2D representation from the side.

HEIGHT:  1    2    3   -3   -2   -1
__   __
__  |__| |__|  __
__  |__| |__|      |__|  __
|__| |__| |__|           |__|


## Test Cases

If you want to try out your solution on a few more inputs, I've put together a couple of test cases here.

## Winning

This is , so the shortest submission (in bytes) wins.

• Ohh boy, get ready for 300+ byte solutions. Good challenge. +1 – totallyhuman Jan 12 '18 at 20:11
• @totallyhuman Nah, Dennis is going to have a 9-byte solution for this in 20 minutes. – Deacon Jan 12 '18 at 20:50
• Does the perspective have to be as shown with the bottom left of the input data in the foreground? The fact that this is not the first or last element of data makes it harder. Is it acceptable to either 1. keep the mapping as is and draw output with the bottom right column in the foreground or 2. draw a mirror image or 90 deg rotation of the data? Either of these would make the last data element correspond with the column in the foreground, which would be easier. – Level River St Jan 12 '18 at 21:41
• I feel inclined to use a real game engine (or part of it) to render a photo and convert it to ASCII – Stan Strum Jan 12 '18 at 22:16
• @LevelRiverSt That seems like a reasonable request - you can choose the order of the 121 input elements to be whatever makes the most sense for your solution, as long as your ordering is consistent. It must be possible to produce every kind of layout that can be produced with the default order. – James Holderness Jan 12 '18 at 22:24

# Charcoal, 7069 68 bytes

≔Ｅ¹¹⮌Ｉ⪪Ｓ,θＦ²Ｆ¹¹Ｆ¹¹Ｆ¹¹«Ｊ⁻⁻⁺λκ×μ³ι⁻λκ≔§§θλμη¿∨⁼±η⊕κ‹κη¿ι“↗⊟&⁹κＵhnI”___


Try it online! Link is to verbose version of code. Explanation:

≔Ｅ¹¹⮌Ｉ⪪Ｓ,θ


Read the array, split each line on commas and cast to integer, but also reverse each line, since we want to draw right-to-left so that left columns overwrite right columns. (Other dimensions already have desired overwriting behaviour.)

Ｆ²Ｆ¹¹Ｆ¹¹Ｆ¹¹«


Loop through i) top lines and bodies k) height l) rows m) columns. (Looping through first top lines and then bodies avoids overwriting bodies with top lines.)

Ｊ⁻⁻⁺λκ×μ³ι⁻λκ


≔§§θλμη


Fetch the height at the current row and column.

¿∨⁼±η⊕κ‹κη


Test whether a cube should be drawn at this height for this row and column.

¿ι“↗⊟&⁹κＵhnI”___


Draw the body or top of the cube.

• When I change the first 3 to a 33, I only get 11 blocks in the tower. In general towers seem to be capped at 11. How does that happen? – Fabian Röling Jan 15 '18 at 13:09
• @Fabian I'm slightly confused that Ｆ¹¹Ｆ¹¹Ｆ¹¹ wasn't a clue... – Neil Jan 15 '18 at 16:06
• I don't know this programming language, I just played around a bit with the TIO link. – Fabian Röling Jan 15 '18 at 20:30

# C,  376  350  313  309  285 bytes

Thanks to @Jonathan Frech for saving four bytes!

#define F for(
char*t,G,*s;i,j,e,k,v,x,y;b(){F s="\\/__//\\__\\ ___ ";*s;--y,s+=5)F e=5;e--;*t=*s<33&*t>32?*t:s[e])t=G[y]+x+e;}f(int*M){F;e<1716;++e)G[e/66][e%66]=32;F k=0;++k<12;)F i=0;i<11;++i)F j=11;j--;v+k||b())x=i+j*3+k,y=14+i-k,(v=M[i*11+j])>=k&&b();F;++e<26;)puts(G+e);}


Try it online!

Unrolled:

#define F for(

char *t, G, *s;
i, j, e, k, v, x, y;

b()
{
F s="\\/__//\\__\\ ___ "; *s; --y, s+=5)
F e=5; e--; *t=*s<33&*t>32?*t:s[e])
t = G[y]+x+e;
}

f(int*M)
{
F; e<1716; ++e)
G[e/66][e%66] = 32;

F k=0; ++k<12;)
F i=0; i<11; ++i)
F j=11; j--; v+k||b())
x = i+j*3+k,
y = 14+i-k,
(v=M[i*11+j])>=k && b();

F; ++e<26;)
puts(G+e);
}

• Can 26*66 not be 1716? – Jonathan Frech Jan 12 '18 at 23:44
• @JonathanFrech Sure, I forgot about that. – Steadybox Jan 12 '18 at 23:44
• *s==32 -> *s<33. – Jonathan Frech Jan 12 '18 at 23:47
• for(e=k=1;e;++k)for(e= -> for(k=1;e;++k)for(e=. – Jonathan Frech Jan 12 '18 at 23:48
• #define B b(...)&++e -> #define B++e&b(...) (assuming b does not depend on e, which I think it does not). – Jonathan Frech Jan 12 '18 at 23:52

# JavaScript (ES6), 277 251 bytes

a=>(n=55,$=f=>[...Array(n)].map((_,i)=>f(i)),S=$(_=>$(_=>' ')),n=11,$(l=>$(z=>$(y=>$(x=>(x=10-x,X=x*3+y+z,Y=y-z+n,Z=a[y][x])<=z&&Z+z+1?0:l?['/\\__\\','\\/__/'].map(s=>S[++Y].splice(X,5,...s)):S[Y].splice(X+1,3,...'___'))))),S.map(r=>r.join).join )  f= a=>(n=55,$=f=>[...Array(n)].map((_,i)=>f(i)),S=$(_=>$(_=>' ')),n=11,$(l=>$(z=>$(y=>$(x=>(x=10-x,X=x*3+y+z,Y=y-z+n,Z=a[y][x])<=z&&Z+z+1?0:l?['/\\__\\','\\/__/'].map(s=>S[++Y].splice(X,5,...s)):S[Y].splice(X+1,3,...'___'))))),S.map(r=>r.join).join
)

console.log(f([
[3, 0, 0, 0, 0, 0, 0, 0, 0, 0, -11],
[2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[1, 0, 0, 7,-7,-7,-7,-7, 7, 0, 0],
[0, 0, 0, 7,-7,-7,-7,-7, 7, 0, 0],
[0, 0, 0, 6, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 5, 0, 0, 0, 0, 0, 0, 0],
[1, 0, 0, 4, 3, 2, 1, 0, 0, 0, 1],
[1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1],
[1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1],
[1, 1, 1, 1, 0, 0, 0, 1, 1, 1, 11]
]))

Saved 26 bytes from @Neil's suggestion.

### Ungolfed

a=>(
n=55,
$=f=>[...Array(n)].map((_,i)=>f(i)), S=$(_=>$(_=>' ')), n=11,$(l=>
$(z=>$(y=>$(x=>( x=10-x, X=x*3+y+z, Y=y-z+n, Z=a[y][x], Z<=z && Z+z+1 || ( l ? ['/\\__\\','\\/__/'].map(s=>S[++Y].splice(X,5,...s)) : S[Y].splice(X+1,3,...'___') ) )))) ), S.map(r=>r.join).join\n )  • ,$(w=>$(z=>$(y=>$(x=>(Z=a[y][x=10-x,X=x*3+y+z,Y=y-z+n,x])<=z&&Z+z+1?0:w?['/\\__\\','\\/__/'].map(s=>S[++Y].splice(X,5,...s)):S[Y].splice(X+1,3,...'___'))))), seems to save 26 bytes. – Neil Jan 13 '18 at 11:26 • @Neil Brilliant! Drawing all the top lines first saves me the trouble of checking for non-spaces. – darrylyeo Jan 13 '18 at 16:14 # Python 2, 243 bytes a=input() s=eval([[' ']*55]*23) for h in range(7986): k=h%3;x=h/3%11;y=h/33%11;z=h/363%11;i=h/3993;u=y+z-x*3+30;v=y-z+10 if~-(z>=a[y][10-x]!=~z): if i*k:s[v+k][u:u+5]='\//\____/\\'[k%2::2] if~-i:s[v][u+1+k]='_' for l in s:print''.join(l)  Try it online! A Python translation of Neil’s Charcoal approach. • Nice to see a golfed Python solution for this. My Python proof-of-concept was more than 900 bytes! – James Holderness Jan 13 '18 at 14:34 • +1+k -> -~k. – Jonathan Frech Jan 13 '18 at 16:42 # APL (Dyalog Unicode), 117116 112 bytesSBCS a←23 55⍴0⋄{i←11 0+⍵+.×3 2⍴3-7897⊤⍨6⍴5⋄(,2 5↑i↓a)←745366⊤⍨10⍴4⋄(3⍴i↓1⌽¯1⊖a)⌈←1}¨j[⍋-/↑j←⍸↑⎕↑¨⊂⌽1,11/⍳2]⋄' _/\'[a]  Try it online! # Tcl, 380 409 bytes User sergiol has been busy crunching this down very nicely: set X [read stdin] proc L {a b c d e s} {time {incr z set y -1 time {incr y set x -1 time {if {abs([set Z [lindex$::X [expr ($y+1)*11-[incr x]-1]]])==$z|$z<$Z} {set s [string repl [string repl $s [set i [expr -3*$x+57*$y-55*abs($z)+701]] $i+$b $a] [incr i$c] $i+$e $d]}} 11} 11} 12 set s} puts [L /\\__\\ 4 56 \\/__/ 4 [L "" -1 -55 ___ 2 [string repe [string repe \ 55]\n 23]]]  Try it online! ### Original Content set xs [read stdin] proc L {a b c d e s} {set z 0 while {[incr z]<12} {set y -1 while {[incr y]<11} {set x -1 while {[incr x]<11} {set Z [lindex$::xs [expr ($y+1)*11-$x-1]]
if {abs($Z)==$z||$z<$Z} {set i [expr -3*$x+57*$y-55*abs($z)+701] set s [string repl [string repl$s $i$i+$b$a] [incr i $c]$i+$e$d]}}}}
set s}
puts [L /\\__\\ 4 56 \\/__/ 4 [L "" -1 -55 ___ 2 [string repe [string repe \  55]\n 23]]]


Try it online!

Alas, it is what it is. It’s only a tad easier on the eyes when “ungolfed”

set s [string repeat [string repeat " " 55]\n 23]

proc loops {s0 i0 io s1 i1} {
set z  0; while {[incr z] < 12} {
set y -1; while {[incr y] < 11} {
set x -1; while {[incr x] < 11} {
set Z [lindex $::xs [expr {($y+1) * 11 - $x - 1}]] if {abs($Z) == $z ||$z < $Z} { set i [expr {-3*$x + 57*$y - 55*abs($z) + 701}]
set ::s [string replace $::s$i $i+$i0 $s0] incr i$io
set ::s [string replace $::s$i $i+$i1 $s1] } } } } } loops "" -1 -55 \ ___ 2 loops /\\__\\ 4 56 \ \\/__/ 4 puts$s


Builds a string, as per requirements. Takes the array from stdin. Goes from bottom to top, front to back, right to left over the string data. Does it in two passes, once for the top edge and again for the rest of the body of each cube.

I tried to make it smaller using some sweet functional lambda mojo, but alas, that made it larger.