I suppose it's not too bad.

    my_nextfloat(x)=with_bigfloat_precision(()->nextfloat(x),x.prec)

Reply via email to