I have written a computer program which solves Sudoku puzzles

[Ecthy]

Dead brains? It's not like they're going to ****** you.

OF course it also depends on how you do them. I developed a method, then they got really boring.

[Colon™]

Sudoku is part of Japan's cultural imperialism.

[Imran Siddiqui]

We must pass a law forcing newspapers to equate at least 20% of their puzzle pages to American games

[Colon™]

We can't have those crazy Japs destroying our heritage.

[Urban Ranger]

I was going to do something like that, only in another language.

[Sava]

I'm going to write a computer program that creates Sudoku puzzles that contain sublimal messages that program people to KILL!

KILLKILLKILLKILLKILLKILLKILLKILLKILLKILLKILLKILLKI LLKILL

[Colon™]

Hopefully "kill" includes suicide, and you'll test the progrom on yourself first.

[notyoueither]

Hmmmm, brains.

[Sava]

Originally posted by Colon
Hopefully "kill" includes suicide, and you'll test the progrom on yourself first.

no...

the only suiciding I will do is I'll suicide bomb your house while you sleep

but after my Sudoku puzzles have caused THE PUBLIC to KILLKILLKILLKILLKILLKILLKILLKILLKILLKILLKILLKILLKI LLKILLKILLKILLKILL

that is of course, if you aren't a victim of a CRAZED SUDOKU PROGRAMMED KILLING MACHINE



MUHAUHUAHUAHHAUHHAHHAHAH

[KrazyHorse]

Originally posted by Imran Siddiqui
Jesus... and I thought I needed a girlfriend

I have one

[Colon™]

You sure?

[VetLegion]

Her name is Compiler. Sometime he cheats her with Interpreter

[KrazyHorse]

Originally posted by Colon
You sure?

I just put it on her 10 minutes ago, so yes.

[Asher]

Not very pretty code. For shame.

Looks like another graph coloring implementation. Go on, be creative!

[KrazyHorse]

Eh. I'm not a computer scientist or even a computer engineer. It works (in all cases I've tested), and that's what counts.

Took out that ugly sequence of inputted values and made it dependent on an external file which has the same format as the puzzle. Blanks (spaces with no preassigned value) should be entered as 0s. The output is now also written to a file, and is in the same intuitive format.

Tried to make a standalone but IDL is acting *****y about doing that.

function square, r, c
s = 3*floor(r/3.0) + floor(c/3.0)
return, s
end

function delta, a, b
s = 0
if (a EQ b) then begin
	s = 1
endif
return, s
end

function eliminator, poss
a = 0L
while a LE 100L do begin
	totalposs0 = total(poss)
	sumvals = total(poss, 3)
	for r = 0, 8 do begin
		for c = 0, 8 do begin
			if (sumvals[r, c] EQ 0) then begin
				poss = fltarr(9, 9, 9)
				a = a + 10000L
			endif
			if (sumvals[r, c] EQ 1) then begin
				for j = 0, 8 do begin
					if (poss[r, c, j] EQ 1) then begin
						for rprime = 0, 8 do begin
							poss[rprime, c, j] = delta(r, rprime)
						endfor
						for cprime = 0, 8 do begin
							poss[r, cprime, j] = delta(c, cprime)
						endfor
						for rprime = 0, 8 do begin
							for cprime = 0, 8 do begin
								if (square(r, c) EQ square(rprime, cprime)) then begin
									poss[rprime, cprime, j] = poss[rprime, cprime, j] * delta(r, rprime) * delta(c, cprime)
								endif
							endfor
						endfor
					endif
				endfor
			endif
		endfor
	endfor
	sumrows = total(poss, 1)
	for c = 0, 8 do begin
		for j = 0, 8 do begin
			if (sumrows[c, j] EQ 0) then begin
				poss = fltarr(9, 9, 9)
				a = a + 10000L
			endif
			if (sumrows[c, j] EQ 1) then begin
				for r = 0, 8 do begin
					if (poss[r, c, j] EQ 1) then begin
						for jprime = 0, 8 do begin
							poss[r, c, jprime] = delta(j, jprime)
						endfor
					endif
				endfor
			endif
		endfor
	endfor
	sumcols = total(poss, 2)
	for r = 0, 8 do begin
		for j = 0, 8 do begin
			if (sumcols[r, j] EQ 0) then begin
				poss = fltarr(9, 9, 9)
				a = a + 10000L
			endif
			if (sumcols[r, j] EQ 1) then begin
				for c = 0, 8 do begin
					if (poss[r, c, j] EQ 1) then begin
						for jprime = 0, 8 do begin
							poss[r, c, jprime] = delta(j, jprime)
						endfor
					endif
				endfor
			endif
		endfor
	endfor
	sumsqrs = fltarr(9, 9)
	for j = 0, 8 do begin
		for r = 0, 8 do begin
			for c = 0, 8 do begin
				sqr = square(r, c)
				sumsqrs(sqr, j) = sumsqrs(sqr, j) + poss[r, c, j]
			endfor
		endfor
	endfor
	for sqr = 0, 8 do begin
		for j = 0, 8 do begin
			if (sumsqrs[sqr, j] EQ 0) then begin
				poss = fltarr(9, 9, 9)
				a = a + 10000L
			endif
			if (sumsqrs[sqr, j] EQ 1) then begin
				for r = 0, 8 do begin
					for c = 0, 8 do begin
						if (square(r, c) EQ sqr) then begin
							if (poss[r, c, j] EQ 1) then begin
								for jprime = 0, 8 do begin
									poss[r, c, jprime] = delta(j, jprime)
								endfor
							endif
						endif
					endfor
				endfor
			endif
		endfor
	endfor
	totalposs1 = total(poss)
	if (totalposs1 EQ totalposs0) then begin
		a = a + 10000L
	endif
	a = a + 1
endwhile
return, poss
end

pro solver
invalues = lonarr(9)
openr, lun, 'invalues.txt', /get_lun
readf, lun, invalues
free_lun, lun
gpos = lonarr(81)
gvalprime = lonarr(81)
for i = 0, 8 do begin
	for j = 0, 8 do begin
		gvalprime[9*i + j] = (floor(invalues[i] / (10L^(8 - j)))) MOD 10L
		if (gvalprime[9*i + j] GE 1) then begin
			gpos[9*i + j] = 1
		endif
	endfor
endfor
n = total(gpos)
;n is the number of values pre-determined by the person who wrote the puzzle
grow = intarr(n)
gcol = intarr(n)
gval = intarr(n)
;given values for given row and given column
;note that row and column numbering begins at 0, but gval must be an integer from 1 to 9
i = 0
j = 0
while (i LT n) do begin
	if (gpos[j] EQ 1) then begin
		grow[i] = floor(j / 9.0)
		gcol[i] = j MOD 9
		gval[i] = gvalprime[j]
		i = i + 1
	endif
	j = j + 1
endwhile
poss = intarr(9, 9, 9)
for r = 0, 8 do begin
	for c = 0, 8 do begin
		for j = 0, 8 do begin
			poss[r, c, j] = 1
		endfor
	endfor
endfor
for i = 0, (n - 1) do begin
	for j = 0, 8 do begin
		if (j NE (gval[i] - 1)) then begin
			poss[grow[i], gcol[i], j] = 0
		endif
	endfor
endfor
poss = eliminator(poss)
if (total(poss) GT 81) then begin
	a = 0L
	b = 0L
	while (a LE 100L) do begin
		sumvals = total(poss, 3)
		possprime = poss
		r = floor(b / 81.0)
		c = (floor(b / 9.0)) MOD 9
		j = b MOD 9
;		print, r
;		print, c
;		print, j
		if ((sumvals[r, c] GE 2) AND (poss[r, c, j] EQ 1)) then begin
;			print, "made it into if loop"
;			print, r
;			print, c
;			print, (j + 1)
			for jprime = 0, 8 do begin
				possprime[r, c, jprime] = delta(j, jprime)
			endfor
			possprime = eliminator(possprime)
			if (total(possprime) EQ 0) then begin
;				print, "Possibility Eliminated!"
				poss[r, c, j] = 0
				poss = eliminator(poss)
			endif
			a = a + 1L
		endif
		if ((total(poss) EQ 81) OR (b GT 729L)) then begin
			a = a + 10000L
		endif
		b = b + 1L
	endwhile
endif
openw, lun, 'solution.txt', /get_lun
if (total(poss) GT 81) then begin
	printf, lun, "fail to completely solve; partial solution begins"
	for r = 0L, 8L do begin
		printf, lun, "beginning of row"
		for c = 0L, 8L do begin
			count = 0L
			p = 0L
			for j = 0, 8 do begin
				p = p + (j + 1) * poss[r, c, j] * (10L^(count))
				count = count + poss[r, c, j]
			endfor
			printf, lun, p
		endfor
	endfor
endif
if (total(poss) LE 81) then begin
	p = lonarr(9)
	for r = 0, 8 do begin
		for c = 0, 8 do begin
			for j = 0, 8 do begin
				p[r] = p[r] + poss[r, c, j] * (j + 1) * (10L^(8 - c))
			endfor
		endfor
		printf, lun, p[r]
	endfor
endif
free_lun, lun
end