float_int

Summary. Convert an XFL to a plain integer, scaled by a number of decimal places.

Signature.

int64_t float_int(int64_t float1, uint32_t decimal_places, uint32_t abs);

Parameters.

Name	Type	Description
float1	int64_t	The XFL to convert.
decimal_places	uint32_t	Number of fractional digits to keep (i.e. scale by 10^decimal_places); 0..15.
abs	uint32_t	> 0 allows negative inputs by returning their absolute value; 0 rejects negatives.

Return value. The integer value of float1 * 10^decimal_places, truncated. float1 == 0 returns 0. A value too small to show at the requested scale returns 0. Errors: INVALID_FLOAT (-10024) if float1 is invalid; INVALID_ARGUMENT (-7) if decimal_places > 15; CANT_RETURN_NEGATIVE (-33) if float1 is negative and abs == 0; TOO_BIG (-3) if the scaled value would overflow an integer.

Common failure patterns.

• Asking for more precision than a value has → 0 (not an error) for values too small; but scaling a large value up so it overflows → TOO_BIG.
• A negative XFL with abs == 0 → CANT_RETURN_NEGATIVE.
• decimal_places > 15 → INVALID_ARGUMENT.

Caveats / notes.

• Use decimal_places == 6 to turn an XFL number of XAH into drops (1 XAH = 1,000,000 drops).
• Truncates toward zero; it does not round.

Minimal example.

int64_t drops = float_int(amount_xah, 6, 0);   // XAH -> drops

Practical example.

ASSERT(float_int(-1, 0, 0)            == INVALID_FLOAT);        // not a valid XFL
ASSERT(float_int(float_one(), 0, 0)   == 1LL);                  // 1.0 -> 1
ASSERT(float_int(float_one(), 15, 0)  == 1000000000000000LL);   // scaled by 10^15
ASSERT(float_int(float_one(), 16, 0)  == INVALID_ARGUMENT);     // decimal_places > 15

Related APIs. float_mantissa, float_set, AMOUNT_TO_DROPS.