adaptable-a-star
Summary
Adaptable and efficient A* graph search and pathfinding implementation that operates on every collection that provides forward iteration (such as std::deque, std::list, custom collections, you name it).
Thanks to the templated design it's suited to work with every graph with positive weighted nodes or edges, little work required.
Usage
Include AStar.hpp and create a specialization of AStar::NodeBase.
For instance, if you choose to store your nodes in a std::deque, the specialization could look like the following:
class MyNode: public AStar<MyNode, std::deque>::NodeBase {
public:
const double distance(const MyNode *rhs) const { /*...*/ }
const double heuristic(const MyNode *rhs) const { /*...*/ }
const Collection successors(const Iterator &collection_begin, const Iterator &collection_end) const { /*...*/ }
const bool operator==(const MyNode *rhs) const { /*...*/ }
}
To perform a search, simply instantiate AStar with the iterators collection.begin, collection.end, path_begin and path_end. The AStar object then holds the optimal path, ready to be traversed.
std::deque<MyNode*> nodes;
AStar<MyNode, std::deque> myAStar(nodes.begin(), nodes.end(), path_begin, path_end);
A reference implementation using a two-dimensional, rectangular, evenly distributed grid (or, with other words, a simple Tile Collision Map) with extensive documentation can be found in Demo.cpp.
Compilation
Compiling the source is easy. Use one of the following console commands...
clang++ -std=c++11 -stdlib=libc++ Demo.cpp -o astar_demo
g++ -std=c++11 Demo.cpp -o astar_demo
..., include AStar.hpp in your existing project, or utilize your favourite IDE. (No, DevC++ is not an IDE you should use.)
License
This software is licensed under the zlib/libpng license. It's free to use this software in non-commercial as well as in commercial projects.
See LICENSE in the project root directory for details.