gloryfish / lua-astar

A simple implementation of the A* (A Star) pathfinding algorithm for lua

Lua A*

About

This is a simple implementation of the A* (A Star) pathfinding algorithm for lua. It is intended to be relatively easy to read and learn from.

It is based on the similarly easy to understand Python implementation by John Eriksson.

http://www.pygame.org/project-AStar-195-.html

Usage

Add astar.lua and middleclass.lua to your project.

Begin by implementing a map handler for your specific game. The map handler's job is to determine which nodes are valid for a given path and to calculate a heuristic score for each.

You can find an example handler in "examplehandler.lua". Simply implement those methods in a table (most likely in your level or game object) and pass that table to the AStar class.

local maphandler = MyMapHandler()
local astar = AStar(maphandler)

When you want to calculate a path

local path = astar:findPath(start, goal) 

"start" and "goal" are locations in your game. In the example handler, they are simple vectors with x, y components. However, they could be anything: hex tiles, states in a game of tic-tac-toe, etc. You simply need to implement the handler methods for your specific game needs.

astar:findPath() will return a Path object with two methods.

path:getTotalMoveCost() returns the total cost of the path. You could use it to limit enemy movement if the path cost was above a certain threshold.

path:getNodes() returns a table containing each node in the path, from start to goal.

Example

Run test.lua to see an example using the TiledMapHandler. This is similar to the implementation used in my Reddit Game Jam 05 submission "Unrequited".

The source code is on GitHub.

Contributors

Sandy Maguire

License

See license.txt

lua-astar's Issues

getBestOpenNode() is O(n)

Replace the linear table lookup in getBestOpenNode with a MinHeap pop.

Need to implement a min heap. Replace astar.on with the heap. Check to see if astar.o needs to be replaced.