Enum FloatValue 

pub enum FloatValue {
    Float {
        value: f64,
        precision: u32,
    },
    Decimal {
        value: BigDecimal,
        precision: u32,
    },
}

The value handed to the float path, mirroring what Python’s dispatcher saw.

base.float2tuple branches on isinstance(value, Decimal): the Decimal arm is exact arbitrary precision, the float arm is lossy f64. They are not interchangeable — issue #603 exists precisely because a float() cast silently rounded 98_746_251_323_029.99 at trillion scale.

Variants

Float

Python passed a float. precision is derived from repr on the Python side.

Fields

value: f64

precision: u32

Decimal

Python passed a Decimal. precision is abs(exponent).

Fields

value: BigDecimal

precision: u32

Implementations

impl FloatValue

pub fn precision(&self) -> u32

pub fn is_negative(&self) -> bool

pub fn has_visible_point(&self) -> bool

Whether Python’s string form of the value shows a decimal point — the "." in str(value) test many self-contained converters route on. repr(float) has a point unless repr picked exponent form (1e+16); str(Decimal) has one exactly when the scale is positive.

pub fn as_whole_int(&self) -> Option<BigInt>

Python’s int(value) == value routing test: Some(int) when the value is whole, None when fractional. NaN/±inf return None — the dispatcher keeps those on the Python side (base’s assert raises OverflowError/ValueError there, outside the whole-value route).

pub fn whole_float_int(&self) -> Option<BigInt>

PR savoirfairelinux/num2words#666: the dispatcher’s isinstance(number, float) and number == int(number) test — Some(int) only for a genuine float (not Decimal) whose value is integral. Routes a whole float through the integer to_ordinal_num so the result formats without the decimal point (“1st”, not “1.0st”). Decimals keep their scale (“5.00th”), matching isinstance(number, float) == False.

Trait Implementations

impl Clone for FloatValue

fn clone(&self) -> Self

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for FloatValue

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.