result = Sar(a, x[, bignum])
a by x bits to the right padding the
holes with the most significant bit of a. This is called an arithmetic shift.
a is converted to a signed 32-bit integer variable before the shift
(unless bignum is set to True).
Starting with Hollywood 9.0, there is an optional bignum argument. If this is
set to True, Sar() will be able to operate on integers larger than 2^31 but keep
in mind that Sar() still won't be possible to use the full 64-bit integer range because
Hollywood's numeric type is a 64-bit floating point number and is thus limited
to integers in the range of [-9007199254740992,9007199254740992].
False) (V9.0)a = Sar(-256, 3)This returns -32.