I'm currently playing with some maze generation algorithms for my logic/puzzle game pack, didn't see any such topic here on forum, so I thought that sharing this could be a good idea. Some maze generation examples available on web have nice visualization, just hope that my would be sufficient to show how each algorithm work. Project includes implementation of several most common maze generation algorithms like: Binary Tree, Sidewinder, Hunt&Kill, DFS/Backtracker, Eller, Prim, Growing Tree. Some were modified to add difficulty variation - mostly length of walk in current direction.
One major addition would be bridges (paths above/beneath another), those simple algorithms have limited possibilities for that, and needs to be used in combination or as addition to some other techniques - like regions, pre-calculating bridge positions and so on. Last algorithm - Growing Tree has basic bridge logic implemented.
One thing that code can be a little messy, most time I use Javascript for such prototyping and code was ported to GLB to share on forum (GLB is also target/final language). Feel free to modify, and new functions.
Controls:
-+ -> switch algorithm
[] -> change map width
<> -> change map height
space -> generate maze
escape -> end