20
\$\begingroup\$

Challenge

Write a program/function that accepts an "image" and outputs a picture maze formed from that image.

Input

Your program should accept two arguments:

  • I, the image to form the maze from
  • S, a boolean specifying whether or not to display the solution to the maze

I is given in the following form:

.......
.#####.
.#####.
#######
.#####.
.#####.
.......

where #'s are cells to be included in the solution path and .'s are cells to be excluded. You may swap out the .'s, #'s and newlines with any character of your choosing as long as they differ from each other. Alternatively, you may accept an actual bitmap of the input image.

Output

Your resulting maze should be in the following form:

###############
#             #
# ### ####### #
# #.........# #
# #.#######.# #
# #.#.......# #
###.#.#########
....#.#........
#####.#.#######
#  ...#.....  #
# #.#######.# #
# #.........# #
# ####### ### #
#   #       # #
###############

where #'s denote walls, .'s denote portions of the path that are part of the solution, and spaces are paths excluded from the solution. The .'s may be replaced by spaces if S is false. Again, characters may be swaped with other characters of your choosing or you may output an actual bitmap of the maze with the solution highlighted.

Additional Details

  • Paths must be one cell wide (cannot have giant pool of empty space be the path)
  • The maze must not contain any loops
  • The maze must be fully connected (all cells must be reachable from the entrance/exit)
  • The maze must be surrounded by walls (unless its an entrance/exit)
  • The solution path must not include dead-ends
  • There must be exactly 1 entrance and 1 exit for the maze
  • The entrance and exit must be aligned to the edge of the grid and adjacent to a cell included in the solution path
  • You may choose where the entrance and exit are placed
  • You may assume that a valid path can be formed from the given input image

(Added for clarification) The diagram below shows how the solution path is correlated to the input image:

Input (I): |    Output:            |    Corresponding Cells: 
           |                       |    (@'s denote #'s from I)
           |                       |
.......    |    ###############    |    ###############
.#####.    |    #             #    |    #             #
.#####.    |    # ### ####### #    |    # ### ####### #
#######    |    # #.........# #    |    # #@.@.@.@.@# #
.#####.    |    # #.#######.# #    |    # #.#######.# #
.#####.    |    # #.#.......# #    |    # #@#@.@.@.@# #
.......    |    ###.#.#########    |    ###.#.#########
           |    ....#.#........    |    .@.@#@#@.@.@.@.
           |    #####.#.#######    |    #####.#.#######
           |    #  ...#.....  #    |    #  @.@#@.@.@  #
           |    # #.#######.# #    |    # #.#######.# #
           |    # #.........# #    |    # #@.@.@.@.@# #
           |    # ####### ### #    |    # ####### ### #
           |    #   #       # #    |    #   #       # #
           |    ###############    |    ###############
           |                       |

Test Cases

Watering can example from Wikipedia:

Input:

..................
..................
.......####.......
......##..##......
.....##....##....#
.....#......#...##
.#############.##.
##..############..
#...###########...
#...##########....
#...##########....
#...##########....
#...##########....
....##########....
....##########....
....##########....
..................
..................

Output (S=false):

#####################################
#   #     #   #   #     #           #
# ### ### ### # # ##### ### ### ### #
#     # #   # # #         # #   # # #
# ### # ##### # ########### # ### # #
# # #         #           # # #   # #
# # # ### ##### # ### ### # ### ### #
#   # # #   #   # # #   # #   # #   #
# ### # ##### ##### ### ##### # # ###
# #   #       #   #   #       # # #  
### ####### ### ### # ### ##### ### #
#   #     # #   #   #   # #   # #   #
# ### ##### # ### ####### # # # # # #
# #       #             #   # #   # #
# # ##### ############# ### ### ### #
#   #   #       #     #   # #   # # #
# ### # ####### # ### ### # # ### # #
# # # #         #   # #   #   #     #
# # # ### ######### # # ##### # #####
# #   # # #       # #   #   # # #   #
# ##### # # ##### # ##### # # ### # #
#       # #   #   # #     # #   # # #
# ### ### ### # ### # ##### ####### #
#   # # #     # #   # #       #     #
# # # # ####### # ### # ##### # ### #
  # # # #   #   #     #     # #   # #
### # # # # # ############# # ### # #
#   # # # #   #         #   # #   # #
##### # # ##### ####### # ### ##### #
#     # # #   #       # #   #       #
##### # # # # ####### # ### #########
#     #     #         #       #     #
# ### ######### ############# # #####
# # #   #     # #       #     #     #
# # ######### # ####### ####### ### #
#             #                 #   #
#####################################

Output (S=true):

#####################################
#   #     #   #   #     #           #
# ### ### ### # # ##### ### ### ### #
#     # #   # # #         # #   # # #
# ### # ##### # ########### # ### # #
# # #         #.......    # # #   # #
# # # ### #####.# ###.### # ### ### #
#   # # #   #...# # #...# #   # #   #
# ### # #####.##### ###.##### # # ###
# #   #    ...#   #   #...    # # #..
### #######.### ### # ###.##### ###.#
#   #     #.#   #   #   #.#   # #...#
# ### #####.# ### #######.# # # #.# #
# #.......#.............#...# #...# #
# #.#####.#############.###.###.### #
#...#   #.......#.....#...#.#...# # #
#.### # #######.#.###.###.#.#.### # #
#.# # #  .......#...#.#...#...#     #
#.# # ###.#########.#.#.##### # #####
#.#   # #.#.......#.#...#...# # #   #
#.##### #.#.#####.#.#####.#.# ### # #
#.      #.#...#...#.#.....#.#   # # #
#.### ###.###.#.###.#.#####.####### #
#.  # # #.....#.#...#.#.....  #     #
#.# # # #######.#.###.#.##### # ### #
..# # # #...#...#.....#.....# #   # #
### # # #.#.#.#############.# ### # #
#   # # #.#...#.........#...# #   # #
##### # #.#####.#######.#.### ##### #
#     # #.#...#.......#.#...#       #
##### # #.#.#.#######.#.###.#########
#     #  ...#.........#.....  #     #
# ### ######### ############# # #####
# # #   #     # #       #     #     #
# # ######### # ####### ####### ### #
#             #                 #   #
#####################################

Bitmap example (same maze as above):

Input: Input bitmap Output (S=false): Output bitmap solution unhighlighted Output (S=true): Output bitmap solution highlighted

\$\endgroup\$
  • 4
    \$\begingroup\$ It might just be me, but I don't see the input picture in the output maze. \$\endgroup\$ – Mike Bufardeci Sep 29 '15 at 0:33
  • \$\begingroup\$ @mike-bufardeci Added a diagram showing the input picture in the output maze. Hope that helps! \$\endgroup\$ – Dendrobium Sep 29 '15 at 1:01
  • 2
    \$\begingroup\$ There doesn't seem to be a rule requiring that the maze is connected. Would this be a valid solution? There also doesn't seem to be a rule that the grid must be surrounded by walls (unless every non-wall is considered an entrance or exit). Would this be a valid solution? \$\endgroup\$ – Martin Ender Sep 29 '15 at 6:24
  • 1
    \$\begingroup\$ Also, please add some more test cases. \$\endgroup\$ – flawr Sep 29 '15 at 15:33
  • \$\begingroup\$ @MartinBüttner The maze should be fully connected and surrounded by walls, edited the question clarifying these points. \$\endgroup\$ – Dendrobium Sep 29 '15 at 16:41
10
\$\begingroup\$

Python 3, 1599 Bytes

I found this to be a fun project and very interesting (and somewhat lengthy). When I saw this, I was reminded of the summer I spent exclusively writing and improving a maze generation algorithm and instantly got to work on this.

After a while, I had an initial draft around 6000 bytes long and I spent the next couple of hours condensing it into the following program:

import math,random;R=range;L=len;T=sorted;P=print;N=random.randint
def M(I,S):
 I=I.rsplit('\n');s=[0]*(1+L(I[0])*2);G=[s]
 for i in R(L(I)):
  r=[0]
  for x in I[i]:r+=x,0
  G+=[r];s=[0]*(1+L(I[0])*2);G+=[s]
 c=E(G,L(G[0])-2,-2);G[c][L(G[0])-1]=1;e=[c,L(G[0])-2];c=E(G,1,2);G[c][0]=1;G[c][1]=1;s=[c,1]
 while s!=e:
  o=[];Q(G,s,e,-2,0,o,0);Q(G,s,e,0,2,o,1);Q(G,s,e,2,0,o,2);Q(G,s,e,0,-2,o,3);o=T(o,key=lambda x:(x[2],-x[1]))[0][0]
  if o==0:G[s[0]-1][s[1]]=1;s[0]-=2
  elif o==1:G[s[0]][s[1]+1]=1;s[1]+=2
  elif o==2:G[s[0]+1][s[1]]=1;s[0]+=2
  else:G[s[0]][s[1]-1]=1;s[1]-=2
  G[s[0]][s[1]]=1
 s=0
 while not s:
  r=N(1,(L(G)-1)/2)*2-1;c=N(1,(L(G[0])-1)/2)*2-1
  if G[r][c]in[1,2]:
   o=[];F(G,r-2,c,o,0);F(G,r,c+2,o,1);F(G,r+2,c,o,2);F(G,r,c-2,o,3)
   try:
    if o[0]==0:G[r-1][c]=2;G[r-2][c]=2
    elif o[0]==1:G[r][c+1]=2;G[r][c+2]=2
    elif o[0]==2:G[r+1][c]=2;G[r+2][c]=2
    else:G[r][c-1]=2;G[r][c-2]=2
   except:0
  s=1
  for x in G:
   if'.'in x:s=0;break
 *s,='#  '
 if S:s[1]='.'
 for x in G:
  for y in x:P(s[y],end='')
  P()
def Q(G,s,e,x,y,o,a,n=0):
 c=lambda x,y:G[s[0]+x][s[1]+y]is'#'
 try:
  if c(x,y):
   try:n+=c(2*x,2*y)
   except:0
   try:n+=c(x+abs(x)-2,y+abs(y)-2)
   except:0
   try:n+=c(x-abs(x)+2,y-abs(y)+2)
   except:0
   o+=[[a,math.sqrt((s[0]+x-e[0])**2+(s[1]+y-e[1])**2),n]]
 except:0
def F(G,r,c,o,a):
 try:
  if G[r][c] is'.':o+=[a]
 except:0
def E(G,y,z,d='#'):
 c=[]
 for x in R(1,L(G)-1,2):
  n=0
  try:n+=G[x-2][y]==d
  except:0
  try:n+=G[x+2][y]==d
  except:0
  n+=G[x][y+z]==d
  if G[x][y]==d:c+=[[x,n]]
 if L(c)>1:c=T(c,key=lambda x:x[1])
 return c[0][0]

Which is about as non-sensical to look at as an ascii-art maze is...

It's worth noting that, since the random function is not used until after the correct path has been found, no matter how many times the same input is given, the route from the start to the end will be the same and, while this program does work for the above examples, sometimes it will be unable to find a solution if it 'drives itself into a wall' so to speak.

When running the above examples, it gives this:

>>> M('''.......
.#####.
.#####.
#######
.#####.
.#####.
.......''',True)
###############
# # #   # #   #
# # # ### # # #
# #...#.....# #
# #.#.#.###.###
#  .#.#.#...# #
###.#.#.#.### #
....#.#.#.#....
# ###.#.#.#.###
# #...#.#.#.  #
# #.###.#.#.# #
# #.....#...# #
### ####### # #
#         # # #
###############
>>> 

this:

>>> M('''..................
..................
.......####.......
......##..##......
.....##....##....#
.....#......#...##
.#############.##.
##..############..
#...###########...
#...##########....
#...##########....
#...##########....
#...##########....
....##########....
....##########....
....##########....
..................
..................''',False)
#####################################
# #     #   # # #   # #   # # # # # #
# ### ##### # # # ### # ### # # # # #
#   # # #   #   # # # #   # # #   # #
### # # ### # ### # # # ### # ### # #
# #     #   # #         # # # # # # #
# ### ##### # # ##### ### # # # # # #
# # #   #   #     # #   # # # # # # #
# # # ##### # ##### ### # # # # # # #
# # # #         # # #         #      
# # # # # # ##### # ### # ######### #
# #   # # # #   # # # # # # # # #   #
# # ####### # ### # # ### # # # # # #
#         # #           #   #     # #
### ##### # # ######### ### ### ### #
#     #   # # #   #   #     #   # # #
# ### ### # # # # # # ####### ### # #
#   # #   # # # # # # #   #       # #
# ##### # # # # # # # # # # # ##### #
#   #   # # # # # # # # #   #     # #
# ####### # # # # # # # #############
#   #     # # # # # # #         #   #
# ####### # # # # # # ##### ##### # #
#   #     # # # # # #           # # #
# ### ### # # # # # ######### ### ###
    # #   # # # # #         #   #   #
# ### # # # # # # ######### ##### ###
#   # # # # # # #                 # #
# # # ### # # # ################### #
# # # # # # # #               #     #
# ### # # # # ############# ### ### #
# # # #     #                     # #
# # ##### # # # ##### # # ##### ### #
# # #     # # #     # # #     #   # #
### ##### # ### # # # ##### # ### # #
#         #   # # # #     # #   # # #
#####################################
>>> 

and this:

>>> M('''..................
..................
.......####.......
......##..##......
.....##....##....#
.....#......#...##
.#############.##.
##..############..
#...###########...
#...##########....
#...##########....
#...##########....
#...##########....
....##########....
....##########....
....##########....
..................
..................''',True)
#####################################
#     #     # #   # # # # # # #     #
##### # # ### ### # # # # # # ### # #
# # # # # # # #     # # # # # # # # #
# # # ### # # ##### # # # # # # # ###
#   # #   # # #.......# #     #   # #
### # ### # # #.#####.# # ####### # #
#   # #   # #...#   #...#   # #   # #
### # ### # #.# # ### #.# ### ### # #
# # # # # #...# #   # #...  # #   #..
# # # # # #.# ### #######.### ### #.#
# #     #  .# #   # # #  .    # #...#
# # #######.##### # # ###.##### #.# #
#  .......#.#...........#...# #...# #
###.# ###.#.#.#########.###.# #.### #
#...# #  .#.#.#...#...#.....#...  # #
#.#######.#.#.#.#.#.#.#######.#######
#.    #  .#.#.#.#.#.#.#...#...#     #
#.#######.#.#.#.#.#.#.#.#.#.### #####
#.    #  .#.#.#.#.#.#.#.#...        #
#.#######.#.#.#.#.#.#.#.#############
#.    # #.#.#.#.#.#.#.#.....        #
#.##### #.#.#.#.#.#.#.#####.#########
#.  #    .#.#.#.#.#.#.......  # #   #
#.# # ###.#.#.#.#.#.########### # ###
..# # #  .#.#.#.#.#.........#   # # #
# # #####.#.#.#.#.#########.# ### # #
# # #    .#.#.#.#...........        #
#########.#.#.#.############### #####
#   #    .#.#.#.............# #     #
### # ###.#.#.#############.# ##### #
#     #  ...#...............      # #
##### # # ### # # # # ### # # ##### #
#     # #   # # # # #   # # #   #   #
####### # ### # # # ##### # ####### #
#       # #   # # #     # #       # #
#####################################
>>> 

For anyone who'd like to try running this program themselves, use the command M(Image, Show solution). I would recommend using the triple-quotes to input the image since otherwise there'll be a lot of back slashes or newline characters involved.

\$\endgroup\$
  • 1
    \$\begingroup\$ Accurate nickname :p \$\endgroup\$ – Fatalize Jun 25 '16 at 17:09
  • 1
    \$\begingroup\$ Nice work! Some tips: Use 0 instead of pass, l.append(a);l.append(b) -> l+=a,b, l.append(a) -> l+=[a], it might be worth assigning '#' to a variable, and def E(G,y,z):\n c=[] -> def E(G,y,z,c=[]): \$\endgroup\$ – Loovjo Jun 25 '16 at 21:00
  • 1
    \$\begingroup\$ Also, if G[r][c]==1 or G[r][c]==2: -> if 0<G[r][c]<3:, s=[0]\n for x in R(L(I[0])*2):s+=[0] -> s=[0]*(1+L(I[0])*2) and (I think, haven't tested it though) G=[s] -> *G=s. \$\endgroup\$ – Loovjo Jun 26 '16 at 22:02
  • \$\begingroup\$ @Loovjo Thanks for the advice, the except:0, l+=a,b and s=[0]*(1+L(I[0])*2) really helped. Unfortunately, for whatever reason, assigning c in the function call doesn't reset it over multiple calls which meant it stopped working, G[r][c] can be a string so I can't use < or > on it and the *G=s gave me a syntax error. Still, great advice. \$\endgroup\$ – Anonymous No Lifer Jun 27 '16 at 10:44
  • 1
    \$\begingroup\$ @AnonymousNoLifer No problems. Also, if G[r][c] can be a string, G[r][c]in[1,2] should work. \$\endgroup\$ – Loovjo Jun 27 '16 at 10:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.