This map doesn't seem to work (17 squares)

[Mouradif]

Hi there !

I noticed a weird behaviour. When I first tested your code, I ran examples in the fmaps/ folder and saw that you successfully find solutions larger than an 8x8 square but just now I wanted to show a friend how it worked and all and I tried a map containing 17 squares like this one :
##..
##..
....
....

And the program kept stuck more than 10 minutes.

Weirdly enough, it took less than 0.00 secs to solve the 16 squares version (the perfect 8x8 square)

[multisme]

Could you send me the exact map your using...? I made some test and can't manage to reproduce the bug. Moreover, the backtracking algorithm used implies some time gap:
When we compute the therotical size of the map containing the solution we make a simple equation
map_size = sqrt((number of pieces) * 4);
Our algorithm is then gonna try to find solutions whith paramater, and then expand the map size when he can't find some.
this whole process create those time_gaps

[Mouradif]

Like I said, I tried with 16 squares and 17 squares, 16 takes less than 0.01 sec (sorry for my first post: not less than 0.00 that would be nonsense) and 17 squares gets stuck (I haven't waited more than 10 minutes). If you really think it's necessary to paste the maps:

16 Squares:

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

and 17 squares:

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

[multisme]

found the problem, gonna fix it this week when i have more time. Thanks a lot for the head up. How is your fillit going.??

[multisme]

I corrected it by adding a variable that check sequences of similar patterns. the previous algorithm would make unnecessary tries.