Game of Life

Yet another WRAPPED implementation of game of live.

How to use

See a live demo here;

1. Define a pattern as:

   var pat = {
     data: "bo$2bo$3o!",
     rule: "23/3"
   };

   The data field is Run Length Encoded (RLE) (see lifewiki:rle). The rule field is the evolution rule, in the format of S/B, where S stands for stays alive and B for born (see lifewiki:rule). The above is a small glider using Conway's life rule.

2. Define a canvas size:

   var param = {
     width: 32,
     height: 16,
     startx: 3,
     starty: 3
   };

   Since this is a wrapped implementation, i.e. the right-most column is adjacent to the left-most one, first row to the last row, we need to know the size (width and height field) of the entire canvas where the automaton evolves. And also the upper-left position to put the initial pattern (startx and starty field).

3. Initialize the life as:

   var gol = life.wrapped();
   var next = gol.reset(pat, param);

   The .reset method returns the same structure as .next where the born field contains all the living cells in the initial pattern, see below.

4. Get the next generation:

   next = gol.next();

   The next is in the format:

   {
     dead: [[x0, y0], [x1, y1], ...],
     born: [[c0, r0], [c1, r1], ...]
   }

   The dead array contains all the cells that are DEAD in this generation. And the born array contains all the cells that are BORN in this generation;

Life API Summary

• life.wrapped()

  Create a life object.

• .reset(pat, param)

  Set the initial pattern for evolution. Returns the same structure as .next.

• .next()

  Get the next generation. See the above How to use section for more on the returned values.

Implementation

This implementation used the concept of ghost cell, the cells that are 1) not live, 2) adjacent to a live cell. Only the ghost cells might be born the next generation.

I maintain an array of live cells and an array of ghost cells for each generation. With the help of ghost cells, I don't need to scan the whole canvas.