fdigits() implementation

First try:

fdigits( v, c=precision(.), s=1 )= {
   localprec( max( c+s+8, precision(.)+1 ) );
   while( type(v) == "t_CLOSURE", v=v() );
   localprec( max( precision(v+.)+1, precision(.) ) );
   v+=.;
   my( r=digits( v * 10^(c+s-2) \ 1 ) );
   if( #r < c+s-1, r= concat( vector(c+s-1-#r), r ) );
   if( s > #r, s= 0 );
   r= r[ s .. -1 ];
   if( c < #r, r= r[ 1 .. c ] );
   r;
}

Some example runs:

? precision(.)
% 38

? fdigits( 1 - 10^-20, 24 )
% [0, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 0, 0, 0]

? #fdigits( Pi )
% 38

? fdigits( Pi, 10, 761 )
ERROR

? fdigits( ()->Pi, 10, 761 )
% [3, 4, 9, 9, 9, 9, 9, 9, 8, 3]

? fdigits( ()->Pi, 10, 19991 )
% [4, 0, 4, 9, 0, 7, 5, 5, 1, 7]

-- Ruud