Some utilities for our dealings with qnans. More...

Classes
union	_IEEESingle
	Data structure to represent the bit pattern of a 32 Bit IEEE 754 floating-point number. More...

Functions
AI_FORCE_INLINE float	get_qnan ()
	Get a fresh qnan. More...

AI_FORCE_INLINE bool	is_not_qnan (float in)
	Check whether a float is NOT qNaN. More...

AI_FORCE_INLINE bool	is_qnan (float in)
	Check whether a given float is qNaN. More...

AI_FORCE_INLINE bool	is_special_float (float in)
	check whether a float is either NaN or (+/-) INF. More...

Detailed Description

Some utilities for our dealings with qnans.

Note: Some loaders use qnans to mark invalid values tempoarily, also Assimp explicitly enforces undefined normals to be set to qnan. qnan utilities are available in standard libraries (C99 for example) but last time I checked compiler coverage was so bad that I decided to reinvent the wheel.

Function Documentation

AI_FORCE_INLINE float get_qnan ( )

Get a fresh qnan.

AI_FORCE_INLINE bool is_not_qnan ( float in )

Check whether a float is NOT qNaN.

Parameters

in	Input value

AI_FORCE_INLINE bool is_qnan ( float in )

Check whether a given float is qNaN.

Parameters

in	Input value

AI_FORCE_INLINE bool is_special_float ( float in )

check whether a float is either NaN or (+/-) INF.

Denorms return false, they're treated like normal values.

Parameters

in	Input value