generating bitboard masks for move

Question

I'm trying to understand how works bitboard representation in chess programming and I can't find usefull information (or just can't translate it correctly ^^) about one detail. My question is, how to generate automaticly masks for move on every position by every piece. I assume its an matrix where each piece type have define every field he can move from this position (array[5][64] for wP, bP, K, R, N, B). For example for Rook on position below, only allowed positions are:

0 0 1 0 0 0 0 0
0 0 1 0 0 0 0 0
0 0 1 0 0 0 0 0
0 0 1 0 0 0 0 0
1 1 R 1 1 1 1 1
0 0 1 0 0 0 0 0
0 0 1 0 0 0 0 0
0 0 1 0 0 0 0 0

I assume i have to create something like this for each piece type and for every tile it can step but do i have to hardcoded it manually to array or there is some possibility to automatize this process and precompute it after program run?

Answer 1

You could fairly easily precompute all of the bitboards that you need, since the rules of chess are well defined. For example, here is a function (in python) that would compute legal moves for a rook:

import sys

def rook(x, y):
    for i in range (1, 8):
        for j in range (1, 8):
            if x == i or y == j:
                sys.stdout.write("1")
            else:
                sys.stdout.write("0")
        sys.stdout.write("\n")

print "Bit board of legal moves for a rook at 1, 3:"
rook(1, 3)

Instead of printing out the bitboard, you would likely store it in a compact format, such as an array of 64 bit values (since an 8x8 board needs 64 bits for each board).

This is a fairly extreme optimization technique, so the details of its implementation are going to get hairy (and be a pain to debug).

I used the wiki bitboard page as reference.