instructions

My method for A* pushes everything into a one dimensional table
(referred to as an array within the rest of the article),
for better traversal through it when needing to check for a few triggers.
A* was initially designed for an arbitrary data structure, so I tried
to stick with that philosophy as much as possible.

However, because of that philosophy, you will need to deform your map
into the one dimension array, with a few requirements for what each of the
array nodes will need. Included in A* is a helper constructor for the nodes
of the array. It requires four things:

	+ Its own array location
	+ Its heuristic score (hScore) (the demo uses taxicab heuristics)
	+ The array location of its neighbor nodes (packaged inside a table)
	+ The distance to the neighbors (this will be relative to the type
	  of map that you use) (packaged inside a table)

The helper node constructor should throw an error if any of its parameters
are wrong.

The demo flattenMap() function is not optimized, as its simply for demo
purposes only. It will flatten a square map, and give the correct
heuristics and everything, but it recreates the flattened array every time
it is called, which causes significant slowdown when you need to change
the array. Some tips for better flattenMap() functions, as well as
more information about using my A* search, or A* in general,
can be found at:

	http://love2d.org/forums/viewtopic.php?f=5&t=7174