Skip to content

Pathfinder and A* solver (astar or a-star) native extension for Defold Engine build on MicroPather.

Notifications You must be signed in to change notification settings

selimanac/defold-astar

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

41 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

A*

This is a pathfinder and A* solver (astar or a-star) native extension for the Defold Engine, built using MicroPather.

Installation

You can use the A* extension in your own project by adding this project as a Defold library dependency.
Open your game.project file, select Project and add a Dependencies field:

https://github.com/selimanac/defold-astar/archive/master.zip

or you can add stable versions from releases.

Examples

Hexagon example: https://github.com/selimanac/defold-astar-hex-example
Basic tilemap example: https://github.com/selimanac/defold-astar-tilemap-example

Old examples: https://github.com/selimanac/defold-astar-examples

Forum

https://forum.defold.com/t/a-native-extension/

Toss a Coin to Your Witcher

If you find my Defold Extensions useful for your projects, please consider supporting it.
I'd love to hear about your projects! Please share your released projects that use my native extensions. It would be very motivating for me.

Release Notes

1.1.1

  • Minor edge case fix on Micropather core.
  • Minor code refactoring.

1.1.0

This is a major release. Consider it a Beta version. New examples on the way.
Feel free to open an issue if you encounter any problems.

  • Now support odd-r, even-r, odd-q and even-q hexagonal grids.
  • Entities: You can now target entities like potions, chests, or enemies, which were not possible to solve on the map before. You can toggle astar.use_entities() before calling the astar.solve() and astar.solve_near() methods.
  • Partial code refactoring(nearly %30) for better readability, performance improvement(minimal) and Defold compatibility. I'll continue to improve it. Almost everything is now preallocated. You may hit the limits. If so, feel free to open an issue
  • astar.print_map() added. You can print the latest map state for debug purposes.
  • astar.solve() and astar.solve_near() now also returns tile content ID.
  • astar.map_vflip() and astar.map_hflip() added. You can now flip the map.
  • astar.reset() added. It cleans up everything.
  • Optional vertically flip map added to astar.setup().
  • API Documentation updated.

!! BREAKING CHANGE !!

The default value for tables and tile positions starts at 1. For backward compatibility, you can simply toggle it in astar.setup() or by using astar.use_zero(true).

1.0.3

  • No changes. This is the final release before v1.1.

1.0.2

  • astar.set_at and astar.get_at added.
  • astar.toogle_zero added.

1.0.1

Small bug fixes. More details

1.0
initial


API

astar.setup(map_width, map_height, direction, allocate, typical_adjacent, [cache], [use_zero], [map_vflip] )

Initial setup. You have to setup the astar before calling any other methods.

PARAMETERS.

  • map_width (int) - Width of your map. This is generally width of your tilemap.

  • map_height (int) - Height of your map. This is generally width of your tilemap.

  • direction (enum) - Movement direction (astar.DIRECTION_FOUR or astar.DIRECTION_EIGHT)

    astar.DIRECTION_FOUR: On a square grid that allows 4 directions of movement using Manhattan distance
    astar.DIRECTION_EIGHT: On a square grid that allows 8 directions of movement using Euclidean distance

  • allocate (int) - How many states should be internally allocated at a time. This can be hard to get correct. The higher the value, the more memory Patfinder will use.

    • If you have a small map (a few thousand states?) it may make sense to pass in the maximum value. This will cache everything, and MicroPather will only need one main memory allocation. For a chess board, allocate would be set to 8x8 (64)
    • If your map is large, something like 1/4 the number of possible states is good.
    • If your state space is huge, use a multiple (5-10x) of the normal path. "Occasionally" call astar.reset_cache() to free unused memory.
  • typical_adjacent (int) - Used to determine cache size. The typical number of adjacent states to a given state. (On a chessboard, 8.) Higher values use a little more memory.

  • cache (bool)[optional] - Turn on path caching. Uses more memory (yet again) but at a huge speed advantage if you may call the pather with the same path or sub-path, which is common for pathing over maps in games. Default is true

  • use_zero (bool)[optional] - Toggle start index 0 or 1 for tables and tile positions. Also you can set it by call astar.use_zero(). Default is false

  • map_vflip (bool)[optional] - Flips the map vertically. This doesn't flip the coordinates. Also you can set it by call astar.map_vflip(). Default is false

EXAMPLE

local map_width = 5
local map_height = 4
local direction = astar.DIRECTION_EIGHT
local allocate = map_width * map_height
local typical_adjacent = 8
local cache = true     -- Optional. Default is true
local use_zero = false -- Optional. Default is false = 1 based
local flip_map = false -- Optional. Default is false

astar.setup(map_width, map_height, direction, allocate, typical_adjacent, cache, use_zero, flip_map)

astar.use_zero(toggle)

Toggle start index 0 or 1 for tables and tile positions.

If set to false, astar.solve, astar.solve_near, astar.get_at, astar.set_at methods expect positions start with 1 and returns table indexes from 1.

Default is false = 1

PARAMETERS

  • toggle (bool) - true/false

EXAMPLE

astar.use_zero(false)

astar.set_map_type(type)

You can set different type of map coordinates. Default is astar.GRID_CLASSIC.

More information about hexagonal grid offset coordinates systems: Red Blob Games

PARAMETERS

  • type (enum) - Type of the map coordinates.

    astar.GRID_CLASSIC : classic grid.
    astar.HEX_ODDR : odd-r hexagonal grid.
    astar.HEX_EVENR : even-r hexagonal grid.
    astar.HEX_ODDQ : odd-q hexagonal grid.
    astar.HEX_EVENQ : even-q hexagonal grid.

EXAMPLE

astar.set_map_type(astar.HEX_ODDR)

astar.set_map(world)

Set your map data.
0.0 or 1.1 is top-left.

*Setting new map data reset the current cache.

PARAMETERS

  • world (table) - Your tilemap data. Keep it simple as much as you can.

EXAMPLE

local world = {
	2, 0, 0, 0, 0,
	0, 0, 0, 0, 0,
	0, 0, 1, 0, 0,
	1, 0, 0, 0, 2,
}

astar.set_map(world)

astar.map_vflip()

Every time you call this function, it flips the map vertically. This does not flip the coordinates

*Flipping the map reset the current cache.

EXAMPLE

local world = {
	2, 0, 0, 0, 0,
	0, 0, 0, 0, 0,
	0, 0, 1, 0, 0,
	1, 0, 0, 0, 2,
}

astar.set_map(world)
astar.map_vflip()
astar.print_map()

astar.map_hflip()

Every time you call this function, it flips the map horizontally. This does not flip the coordinates

*Flipping the map reset the current cache.

EXAMPLE

local world = {
	2, 0, 0, 0, 0,
	0, 0, 0, 0, 0,
	0, 0, 1, 0, 0,
	1, 0, 0, 0, 2,
}

astar.set_map(world)
astar.map_hflip()
astar.print_map()

astar.set_costs(costs)

Set costs for your passable tiles on your world table. This table keys determines the passable area. In this example only numbered "0" tiles are passable.

Table's sum must be the astar.DIRECTION_FOUR (4) or astar.DIRECTION_EIGHT (8). In this example we want to move 8 direction.

*Setting new cost data reset the current cache.

PARAMETERS.

  • costs (table) - Your tile cost data.

EXAMPLE

-- For astar.DIRECTION_EIGHT
local costs = {
    [0] = {
        1.0, -- E
        1.0, -- N
        1.0, -- W
        1.0, -- S
        1.41, -- NE
        1.41, -- NW
        1.41, -- SW
        1.41 -- SE
    }
}

astar.set_costs(costs)
-- OR
-- For astar.DIRECTION_FOUR
local costs = {
    [0] = {
        1.0, -- E
        1.0, -- N
        1.0, -- W
        1.0 -- S
    }
}

astar.set_costs(costs)

astar.set_entities(entities)

Entities are usally not passible items like enemies or chests. You might want to target those tiles but also want to keep them not passable. You can set those entity ids as table. Your world table should contain those entities.

You have to toggle astar.get_entities() before calling astar.solve() or astar.solve_near().

PARAMETERS

  • entities (table) - Your entities data.

EXAMPLE

-- IDs of the entities
local entities = {
	1, 2
}

astar.set_entities(entities)

astar.use_entities(toogle).

If you want to include your entities to your solve operations, you have to set it true before calling astar.solve() or astar.solve_near().

If set the true:

astar.solve() only includes the end/last target tile to the path result. Other entities along the path stays as not passable.

astar.solve_near() includes all entities from entities table.

*When you call the astar.use_entities(), the current cache will be reset.

PARAMETERS

  • toogle (boolean) - Toogle true or false. Default is false

EXAMPLE

-- IDs of the entities
local entities = {
	1, 2
}

astar.set_entities(entities)
astar.use_entities(true)

astar.get_at(x, y)

Returns the value from the map array by coordinates.

PARAMETERS

X, Y of the tile possition on array not the screen position.

  • x (int) - Tile X
  • y (int) - Tile Y

RETURN

  • value (int)

EXAMPLE

local value = astar.get_at(1, 1)

print(value)

astar.set_at(x, y, value)

Set your value to the map array by coordinates.

*Setting new data reset the current cache.

PARAMETERS

X, Y of the tile possition on array not the screen position.

  • x (int) - Tile X
  • y (int) - Tile Y
  • value (int)

EXAMPLE

astar.set_at(1, 1, 0) 

astar.print_map()

Prints the map state for debug purposes.

EXAMPLE

astar.print_map()

cols/x:
1	2	3	4	5	
-	-	-	-	-	
2	0	0	0	0		- row/y: 1
0	0	0	0	0		- row/y: 2
0	0	1	0	0		- row/y: 3
1	0	0	0	2		- row/y: 4

astar.solve(start_x, start_y, end_x, end_y)

Solves the path.

PARAMETERS

X, Y of the tile possition on array not the screen position.

  • start_x (int) - Start tile X
  • start_y (int) - Start tile Y
  • end_x (int) - End tile X
  • end_y (int) - End tile Y

RETURN

  • status (enum) -

    astar.SOLVED: Path solved
    astar.NO_SOLUTION: Can't find the path
    astar.START_END_SAME: Start and End is the same

  • size (int) - Size of the path.

  • total_cost (int) - Total cost of the path

  • path (table) - Table with x, y coordinates and tile ID. First value is the given start point.

EXAMPLE

local start_x = 1
local start_y = 1
local end_x = 3
local end_y = 3

local status, size, total_cost, path = astar.solve(start_x, start_y, end_x, end_y)

if status == astar.SOLVED then
	print("SOLVED")
	print("Path Size", size)
	print("Total Cost:", total_cost)

	for _, tile in ipairs(path) do
		print("x:", tile.x, "y: ", tile.y, "tile ID: ", tile.id)
	end
elseif status == astar.NO_SOLUTION then
	print("NO_SOLUTION")
elseif status == astar.START_END_SAME then
	print("START_END_SAME")
end

astar.solve_near(start_x, start_y, max_cost)

Finds the neighbours according to given cost.

PARAMETERS

  • start_x (int) - Start tile X
  • start_y (int) - Start tile Y
  • max_cost (float) - Maximun cost for finding neighbours

RETURN

  • near_status (enum) -

    astar.SOLVED: Path solved
    astar.NO_SOLUTION: Can't find the path
    astar.START_END_SAME: Start and End is the same

  • near_size (enum) - Size of the found neighbours.

  • nears (enum) - Table with x and y coordinates. First value is the given start point.

EXAMPLE

local start_x = 1
local start_y = 1
local max_cost = 3.0 -- near

local near_status, near_size, nears = astar.solve_near(start_x, start_y, max_cost)

if near_status == astar.SOLVED then
	print("NEAR SOLVED")
	print("Near Size:", near_size)
	for _, tile in ipairs(nears) do
		print("x:", tile.x, "y: ", tile.y, "tile ID: ", tile.id)
	end
elseif near_status == astar.NO_SOLUTION then
	print("NO_SOLUTION")
elseif near_status == astar.START_END_SAME then
	print("START_END_SAME")
end

astar.reset_cache()

If your state space is huge, occasionally call astar.reset_cache() to free unused memory.

EXAMPLE

astar.reset_cache()

astar.reset()

This method removes everything releated to astar library and frees the memory. It may be used when the player completes/exits the level/game.

EXAMPLE

astar.reset()

Games using A-Star

Skull Horde

Skull Horde

Developer: 8BitSkull
Publisher: 8BitSkull
Release Date: TBA
Steam: https://store.steampowered.com/app/3199360/Skull_Horde/

Misland

Misland

Developer: Seal Unicron Games
Release Date: 11 Nov, 2023
Poki: https://poki.com/en/g/misland
Google Play: https://play.google.com/store/apps/details?id=com.sealunicorngames.misland

Monkey Mart

Monkey Mart

Developer: Tiny Dobbins
Release Date: 8 Dec, 2022
Poki: https://poki.com/en/g/monkey-mart
Google Play: https://play.google.com/store/apps/details?id=com.tinydobbins.monkeymart
App Store: https://apps.apple.com/tr/app/monkey-mart/id6480208265

Xfire

Xfire

Developer: 171Dev
Release Date: 28 May, 2024
itch.io: https://benjames171.itch.io/xfire
Source: https://github.com/benjames-171/xfire