Commit 96ed2133 authored by Paul Sokolovsky's avatar Paul Sokolovsky
Browse files

objfloat: Quick&dirty implementation of float floor division.

TODO: Likely doesn't match Python semantics for negative numbers.
parent 96eec4f8
......@@ -111,13 +111,15 @@ mp_obj_t mp_obj_float_binary_op(int op, mp_float_t lhs_val, mp_obj_t rhs_in) {
case MP_BINARY_OP_INPLACE_SUBTRACT: lhs_val -= rhs_val; break;
case MP_BINARY_OP_MULTIPLY:
case MP_BINARY_OP_INPLACE_MULTIPLY: lhs_val *= rhs_val; break;
/* TODO floor(?) the value
// TODO: verify that C floor matches Python semantics
case MP_BINARY_OP_FLOOR_DIVIDE:
case MP_BINARY_OP_INPLACE_FLOOR_DIVIDE: val = lhs_val / rhs_val; break;
*/
case MP_BINARY_OP_INPLACE_FLOOR_DIVIDE:
lhs_val = MICROPY_FLOAT_C_FUN(floor)(lhs_val / rhs_val);
goto check_zero_division;
case MP_BINARY_OP_TRUE_DIVIDE:
case MP_BINARY_OP_INPLACE_TRUE_DIVIDE:
lhs_val /= rhs_val;
check_zero_division:
if (isinf(lhs_val)){ // check for division by zero
nlr_jump(mp_obj_new_exception_msg(&mp_type_ZeroDivisionError, "float division by zero"));
}
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment