Author: Robert Hyatt
Date: 17:57:07 04/06/04
Go up one level in this thread
On April 06, 2004 at 18:15:40, Vaclav Visnadjek wrote:
>Last year, the following occurred to me:
>–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
>BITBOARD *bishopmoves(POSITION *posit)
>{
>bishmoves=calloc(64,sizeof(BITBOARD));
>int bishmovsq[28]={7, 14, 21, 28, 35, 42, 49,
> 9, 18, 27, 36, 45, 54, 63,
> -7, -14, -21, -28, -35, -42, -49,
> -9, -18, -27, -36, -45, -54, -63};
>int h;
>int i;
>for (h=0; h<64; h++)
>{
> for (i=0; i<28; i++)
> {
> if (h+bishmovsq[i]>=0 &&
> h+bishmovsq[i]<64 &&
> ((h+bishmovsq[i])/8+(h+bishmovsq[i])%8)%2 == ((h/8)+(h%8))%2)
> {
> bishmoves[h] |= mask[h+bishmovsq[i]];
> if ((mask[h+bishmovsq[i]] & posit->allpieces) != 0)
> {
> i = (7 * (1 + i/7))-1;
> }
> }
> }
>}
>return(bishmoves);
>}
>
>BITBOARD *rookmoves(POSITION *posit)
>{
>rkmoves=calloc(64,sizeof(BITBOARD));
>int rookmovsq[28]={1, 2, 3, 4, 5, 6, 7,
> 8, 16, 24, 32, 40, 48, 56,
> -1, -2, -3, -4, -5, -6, -7,
> -8, -16, -24, -32, -40, -48, -56};
>int h;
>int i;
>for (h=0; h<64; h++)
>{
> for (i=0; i<28; i++)
> {
> if (h+rookmovsq[i]>=0 &&
> h+rookmovsq[i]<64 &&
> ((h+rookmovsq[i])/8 == h/8 || (h+rookmovsq[i])%8 == h%8))
> {
> rkmoves[h] |= mask[h+rookmovsq[i]];
> if ((mask[h+rookmovsq[i]] & posit->allpieces) != 0)
> {
> i = (7 * (1 + i/7))-1;
> }
> }
> }
>}
>return(rkmoves);
>}
>––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
>With all the necessary adjustments, can them be used as an alternative for
>rotated bitboards? If so, could the main idea help to increase the speed of
>calculations?
The point of rotated bitboards is that there are no loops for each direction.
So I don't quite understand your question...
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.