redUniverse: Support for Discrete Worlds

clean-up #30:

As the last thing for this month-of-cleaning-up-old-code-and-taking-care-of-forgotten-projects, I finally wrote some methods that I had been planning for a long time. They add support for discrete worlds to my redUniverse quark.

It's all fairly simple.

In a 2 dimensional world there are 8 neighbouring cells/locations. The surroundings method returns them.

a= RedWorld(RedVector[100, 200])  //a 2D world
a.surroundings
[ [ -1, -1 ], [ -1, 0 ], [ -1, 1 ], [ 0, -1 ], [ 0, 1 ], [ 1, -1 ], [ 1, 0 ], [ 1, 1 ] ]

And in a 3 dimensional world, the number of surrounding cells grows to 26. That is 3 * 3 * 3 - 1 where the minus one is the [0, 0] location.

a= RedWorld(RedVector[100, 200, 300])  //a 3D world
a.surroundings
[ [ -1, -1, -1 ], [ -1, -1, 0 ], [ -1, -1, 1 ], [ -1, 0, -1 ], [ -1, 0, 0 ], [ -1, 0, 1 ], [ -1, 1, -1 ], [ -1, 1, 0 ], [ -1, 1, 1 ], [ 0, -1, -1 ], [ 0, -1, 0 ], [ 0, -1, 1 ], [ 0, 0, -1 ], [ 0, 0, 1 ], [ 0, 1, -1 ], [ 0, 1, 0 ], [ 0, 1, 1 ], [ 1, -1, -1 ], [ 1, -1, 0 ], [ 1, -1, 1 ], [ 1, 0, -1 ], [ 1, 0, 0 ], [ 1, 0, 1 ], [ 1, 1, -1 ], [ 1, 1, 0 ], [ 1, 1, 1 ] ]

And the numbers for 4, 5 and 6 dimensional worlds (not that I ever used >3) are 80, 242, 728 respectively. (A RedWorld can have any number of dimensions.)

Also it is possible to not only get the directly adjacent cell, but neighbours further away. This example bumps up the surroundingArea variable from the default 1 to 2. Now the surroundings are all the cells next to and two steps away from [0, 0].

a= RedWorld(RedVector[100, 200])  //a 2F world
a.surroundingArea= 2
a.surroundings
[ [ -2, -2 ], [ -2, -1 ], [ -2, 0 ], [ -2, 1 ], [ -2, 2 ], [ -1, -2 ], [ -1, -1 ], [ -1, 0 ], [ -1, 1 ], [ -1, 2 ], [ 0, -2 ], [ 0, -1 ], [ 0, 1 ], [ 0, 2 ], [ 1, -2 ], [ 1, -1 ], [ 1, 0 ], [ 1, 1 ], [ 1, 2 ], [ 2, -2 ], [ 2, -1 ], [ 2, 0 ], [ 2, 1 ], [ 2, 2 ] ]

That is 24 neighbour locations per single cell in a 2D world.

So the surroundings method only give relative positions and the size of the neighbourhood. Not so useful. But there are the two other methods called surroundingLocations and neighbours that is what one should use. surroundingLocations takes an object and returns a list of locations depending on the current surroundings.

a= RedWorld(RedVector[100, 200])  //a 2D world
b= RedObject(a, RedVector[10, 20])  //an object at location [10, 20]
a.surroundingLocations(b)  //get the surrounding locations of object b
[ RedVector[ 9, 19 ], RedVector[ 9, 20 ], RedVector[ 9, 21 ], RedVector[ 10, 19 ], RedVector[ 10, 21 ], RedVector[ 11, 19 ], RedVector[ 11, 20 ], RedVector[ 11, 21 ] ]

Last the neighbours method that returns an array of any nearby objects.

a= RedWorld(RedVector[100, 200])  //a 2D world
b= RedObject(a, RedVector[10, 20])  //an object at location [10, 20]
c= RedObject(a, RedVector[11, 21])  //an object at location [11, 21]
a.neighbours(b)  //get the neighbouring objects of object b
[ a RedObject ]

The different worlds deals with border conditions differently. RedWorld wraps all the locations around and RedWorld3 filters out locations. Compare...

a= RedWorld(RedVector[100, 200])  //a 2D world without borders
b= RedObject(a, RedVector[0, 0])  //an object at upper left corner location [0, 0]
a.surroundingLocations(b)  //get the surrounding locations of object b
[ RedVector[ 99, 199 ], RedVector[ 99, 0 ], RedVector[ 99, 1 ], RedVector[ 0, 199 ], RedVector[ 0, 1 ], RedVector[ 1, 199 ], RedVector[ 1, 0 ], RedVector[ 1, 1 ] ]

a= RedWorld3(RedVector[100, 200])  //a 2D world with borders
b= RedObject(a, RedVector[0, 0])  //an object at upper left corner location [0, 0]
a.surroundingLocations(b)  //get the surrounding locations of object b
[ RedVector[ 0, 1 ], RedVector[ 1, 0 ], RedVector[ 1, 1 ] ]

The neighbours method is quite slow at the moment, but I hope to be able to speed it up considerably later on.

Anyway, here is the complete SVN diff: sourceforge.net/p/quarks/code/1765/