#include <evo/type.h>

Inheritance diagram for FloatT< T >:

Detailed Description

template<class T>
struct evo::FloatT< T >

Nullable basic floating-point base type.

Template Parameters

T	Floating-point type

Floating point types:

Float, FloatD, FloatL

Examples – using Float type (and float type) here, other supported floating point types work the same:

Assigned like a number:
Float val(12.3f); // Set to 12.3
val = 1.23f; // Set to 1.23
Can hold null value:
Float val; // Null by default
val = 56.7f; // Set to 56.7, no longer null
val.set(); // Set to null
Evaluate as bool to check non-null and non-zero:
Float val; // Null by default
if (val) { // false
}
if (!val) { // true
}
val = 0.0f;
if (val) { // false
}
val = 1.0f;
if (val) { // true
}
Use denull() to clear null flag and update value via reference – useful for decrement/increment:
Float val; // Null by default
++val.denull(); // Remove null flag (value is 0.0) and increment so value is now 1.0
Dereference to get raw value:
Float val; // Null by default
if (*val == 1.23f) { // Dereference for raw value -- false: 0.0 when null
}
Comparison operator overloads:
Float val; // Null by default, with raw value 0
if (*val == 0.0f) { // true: raw value is 0.0
}
if (val == 0.0f) { // false: null is not 0.0
}
if (val < 0.0f) { // true: null is less than any value
}
if (*val < 0.0f) { // false: raw value is 0.0
}
Some traits and helpers
bool equal = Float::eq(1.23, 1.23); // check whether approximately equal, accounts for floating point rounding errors

Public Types
typedef void	EvoNullableType
	Identify as nullable type. More...

typedef FloatT< T >	This
	This non-POD type More...

typedef T	Type
	Wrapped POD type. More...

Public Member Functions
	FloatT ()
	Constructor. More...

	FloatT (const FloatT< T > &val)
	Copy constructor. More...

	FloatT (T val)
	Constructor. More...

int	compare (const Nullable< T > &val) const
	Comparison. More...

int	compare (T val) const
	Comparison. More...

T &	denull ()
	Clears null flag and returns value reference. More...

bool	eq1 (const Nullable< T > &val)
	Get whether approximately equal to given value. More...

bool	eq1 (T val)
	Get whether approximately equal to given value. More...

bool	null () const
	Get whether null. More...

	operator SafeBoolType () const
	Safe (explicit) evaluation as bool type. More...

bool	operator! () const
	Negation operator returns whether null or 0. More...

bool	operator!= (const Nullable< T > &val) const
	Inequality operator. More...

bool	operator!= (T val) const
	Inequality operator. More...

const T &	operator* () const
	Dereference for explicit conversion to underlying type (const). More...

T &	operator* ()
	Dereference for explicit conversion to underlying type. More...

bool	operator< (const Nullable< T > &val) const
	Less than operator. More...

bool	operator< (T val) const
	Less than operator. More...

bool	operator<= (const Nullable< T > &val) const
	Less than or equal operator. More...

bool	operator<= (T val) const
	Less than or equal operator. More...

FloatT< T > &	operator= (const FloatT< T > &val)
	Assignment operator. More...

FloatT< T > &	operator= (T val)
	Assignment operator. More...

FloatT< T > &	operator= (ValNull)
	Assignment operator to set as null. More...

bool	operator== (const Nullable< T > &val) const
	Equality operator. More...

bool	operator== (T val) const
	Equality operator. More...

bool	operator> (const Nullable< T > &val) const
	Greater than operator. More...

bool	operator> (T val) const
	Greater than operator. More...

bool	operator>= (const Nullable< T > &val) const
	Greater than or equal operator. More...

bool	operator>= (T val) const
	Greater than or equal operator. More...

Nullable< T > &	set ()
	Set as null. More...

Nullable< T > &	set (const Nullable< T > &src)
	Set as copy of given value. More...

Nullable< T > &	set (T src)
	Set as given value. More...

bool	valid () const
	Get whether valid (not null). More...

const T &	value () const
	Get underlying value (const). More...

T &	value ()
	Get underlying value. More...

T	value (T defval) const
	Get underlying value or given default if null. More...

Static Public Member Functions
static T	eps ()
	Get machine epsilon. More...

static bool	eq (T val1, T val2)
	Get whether values are approximately equal. More...

static bool	eq (T val1, T val2, T eps)
	Get whether values are approximately equal using given epsilon value. More...

static T	exp10 (T num, int exp)
	Multiply number by 10 raised to exponent. More...

static T	fexp10 (int &exp, T num)
	Extract normalized base 10 mantissa and exponent from number. More...

static T	inf ()
	Get infinity value. More...

static bool	inf (T num)
	Check whether value is infinite. More...

static T	max ()
	Get maximum normalized value. More...

static int	maxdigits_prec (int exp, int precision)
	Get max formatting digits with given exponent and precision, including sign and any additional chars (used internally). More...

static int	maxexp ()
	Get maximum allowed exponent. More...

static T	min ()
	Get minimum normalized value. More...

static int	minexp ()
	Get minimum allowed exponent. More...

static bool	nan (T num)
	Get whether value is Not-A-Number (NaN). More...

static T	nan ()
	Get Not-A-Number (NaN) value. More...

static T	precision ()
	Get best precision value. More...

Static Public Attributes
static const int	BYTES = sizeof(T)
	Type size in bytes. More...

static const bool	IS = IsFloat<T>::value
	Whether type is really a floating point type. More...

static const int	MAXDIGITS = std::numeric_limits<T>::digits10
	Maximum significant digits without precision loss. More...

static const int	MAXDIGITS_AUTO = MAXDIGITS + 15
	Max formatting digits with auto precision (used internally) More...

static const bool	NANOK = std::numeric_limits<T>::has_quiet_NaN
	Whether Not-A-Number (NaN) is supported. More...

static const bool	SIGN = true
	Whether type is signed. More...

Member Typedef Documentation

◆ EvoNullableType

typedef void EvoNullableType

inherited

Identify as nullable type.

◆ This

typedef FloatT<T> This

This non-POD type

◆ Type

typedef T Type

Wrapped POD type.

Constructor & Destructor Documentation

◆ FloatT() [1/3]

FloatT ( )

inline

Constructor.

◆ FloatT() [2/3]

FloatT ( const FloatT< T > & val )

inline

Copy constructor.

Parameters

val	Value to copy

◆ FloatT() [3/3]

FloatT ( T val )

inline

Constructor.

Parameters

val	Value to set

Member Function Documentation

◆ compare() [1/2]

int compare ( const Nullable< T > & val ) const

inlineinherited

Comparison.

Parameters

val	Value to compare to

Returns: Result (<0 if this is less, 0 if equal, >0 if this is greater)

◆ compare() [2/2]

int compare ( T val ) const

inlineinherited

Comparison.

If null, the result is always -1 as null compares as less than any non-null value

Parameters

val	Value to compare to

Returns: Result (<0 if this is less, 0 if equal, >0 if this is greater)

◆ denull()

T& denull ( )

inlineinherited

Clears null flag and returns value reference.

Value is unchanged, POD types default to 0 when null
Useful for updating the underlying value by reference
Note: Nullable<bool> template specialization returns by value instead of reference

Returns: Whether valid

◆ eps()

static T eps ( )

inlinestatic

Get machine epsilon.

This is the difference between 1 and the least value greater than 1 that is representable.

Returns: Epsilon value

◆ eq() [1/2]

static bool eq	(	T	val1,
		T	val2
	)

inlinestatic

Get whether values are approximately equal.

This uses the system epsilon value for current type to determine whether the values are close enough to be considered equal
For better accuracy use eq(T,T,T) with a more relevant epsilon value

Parameters

val1	First value to compare
val2	Second value to compare

Returns: Whether approximately equal

◆ eq() [2/2]

static bool eq	(	T	val1,
		T	val2,
		T	eps
	)

inlinestatic

Get whether values are approximately equal using given epsilon value.

Parameters

val1	First value to compare
val2	Second value to compare
eps	Epsilon value to use for comparison – values are approximately equal if difference <= eps

Returns: Whether approximately equal

◆ eq1() [1/2]

bool eq1 ( const Nullable< T > & val )

inline

Get whether approximately equal to given value.

This uses the system epsilon value for current type to determine whether the values are close enough to be considered equal

Parameters

val	Value to compare to

Returns: Whether approximately equal

◆ eq1() [2/2]

bool eq1 ( T val )

inline

Get whether approximately equal to given value.

This uses the system epsilon value for current type to determine whether the values are close enough to be considered equal

Parameters

val	Value to compare to

Returns: Whether approximately equal

◆ exp10()

static T exp10	(	T	num,
		int	exp
	)

inlinestatic

Multiply number by 10 raised to exponent.

Parameters

num	Number to multiply
exp	Exponent to use

Returns: Result

◆ fexp10()

static T fexp10	(	int &	exp,
		T	num
	)

inlinestatic

Extract normalized base 10 mantissa and exponent from number.

Parameters

exp	Stores exponent value [out]
num	Number to use

Returns: Mantissa value

◆ inf() [1/2]

static T inf ( )

inlinestatic

Get infinity value.

Returns: Value

◆ inf() [2/2]

static bool inf ( T num )

inlinestatic

Check whether value is infinite.

Returns: Value

◆ max()

static T max ( )

inlinestatic

Get maximum normalized value.

Returns: Value

◆ maxdigits_prec()

static int maxdigits_prec	(	int	exp,
		int	precision
	)

inlinestatic

Get max formatting digits with given exponent and precision, including sign and any additional chars (used internally).

Returns: Max formatting digits

◆ maxexp()

static int maxexp ( )

inlinestatic

Get maximum allowed exponent.

Returns: Value

◆ min()

static T min ( )

inlinestatic

Get minimum normalized value.

Returns: Value

◆ minexp()

static int minexp ( )

inlinestatic

Get minimum allowed exponent.

Returns: Value

◆ nan() [1/2]

static bool nan ( T num )

inlinestatic

Get whether value is Not-A-Number (NaN).

Returns: Whether NaN, always false if NaN not supported

◆ nan() [2/2]

static T nan ( )

inlinestatic

Get Not-A-Number (NaN) value.

This value doesn't work for comparison. Use nan(T) instead

Returns: NaN value, 0.0 if not supported

◆ null()

bool null ( ) const

inlineinherited

Get whether null.

Returns: Whether null

◆ operator SafeBoolType()

operator SafeBoolType ( ) const

inlineinherited

Safe (explicit) evaluation as bool type.

This is called when object is directly evaluated as a bool, and is equivalent to: !operator!()
See SafeBool

◆ operator!()

bool operator! ( ) const

inlineinherited

Negation operator returns whether null or 0.

Returns: Whether null or 0

◆ operator!=() [1/2]

bool operator!= ( const Nullable< T > & val ) const

inlineinherited

Inequality operator.

Parameters

val	Value to compare to

Returns: Whether inequal

◆ operator!=() [2/2]

bool operator!= ( T val ) const

inlineinherited

Inequality operator.

If null, the result is always true as null is not equal to any non-null value

Parameters

val	Value to compare to

Returns: Whether inequal, true if null

◆ operator*() [1/2]

const T& operator* ( ) const

inlineinherited

Dereference for explicit conversion to underlying type (const).

Alternative to value().

Returns: Underlying type value

◆ operator*() [2/2]

T& operator* ( )

inlineinherited

Dereference for explicit conversion to underlying type.

Alternative to value().

Returns: Underlying type value

◆ operator<() [1/2]

bool operator< ( const Nullable< T > & val ) const

inlineinherited

Less than operator.

If null, null is less than any non-null value

Parameters

val	Value to compare to

Returns: Whether less than val

◆ operator<() [2/2]

bool operator< ( T val ) const

inlineinherited

Less than operator.

If null, null is less than any non-null value

Parameters

val	Value to compare to

Returns: Whether less than val

◆ operator<=() [1/2]

bool operator<= ( const Nullable< T > & val ) const

inlineinherited

Less than or equal operator.

If null, null is less than or equal to any value

Parameters

val	Value to compare to

Returns: Whether less than or equal to val, true if null

◆ operator<=() [2/2]

bool operator<= ( T val ) const

inlineinherited

Less than or equal operator.

If null, null is less than or equal to any value

Parameters

val	Value to compare to

Returns: Whether less than or equal to val, true if null

◆ operator=() [1/3]

FloatT<T>& operator= ( const FloatT< T > & val )

inline

Assignment operator.

Parameters

val	Value to copy

Returns: This

◆ operator=() [2/3]

FloatT<T>& operator= ( T val )

inline

Assignment operator.

Parameters

val	Value to set

Returns: This

◆ operator=() [3/3]

FloatT<T>& operator= ( ValNull )

inline

Assignment operator to set as null.

Returns: This

◆ operator==() [1/2]

bool operator== ( const Nullable< T > & val ) const

inlineinherited

Equality operator.

Parameters

val	Value to compare to

Returns: Whether equal

◆ operator==() [2/2]

bool operator== ( T val ) const

inlineinherited

Equality operator.

If null, the result is always false as null is not equal to any non-null value

Parameters

val	Value to compare to

Returns: Whether equal, false if null

◆ operator>() [1/2]

bool operator> ( const Nullable< T > & val ) const

inlineinherited

Greater than operator.

If null, null is not greater than any non-null value

Parameters

val	Value to compare to

Returns: Whether greater than val, false if null

◆ operator>() [2/2]

bool operator> ( T val ) const

inlineinherited

Greater than operator.

If null, null is not greater than any non-null value

Parameters

val	Value to compare to

Returns: Whether greater than val, false if null

◆ operator>=() [1/2]

bool operator>= ( const Nullable< T > & val ) const

inlineinherited

Greater than or equal operator.

If null, null is not greater than any non-null value

Parameters

val	Value to compare to

Returns: Whether greater than or equal to val

◆ operator>=() [2/2]

bool operator>= ( T val ) const

inlineinherited

Greater than or equal operator.

If null, null is not greater than any non-null value

Parameters

val	Value to compare to

Returns: Whether greater than or equal to val

◆ precision()

static T precision ( )

inlinestatic

Get best precision value.

Returns: Value

◆ set() [1/3]

Nullable<T>& set ( )

inlineinherited

Set as null.

Returns: This

◆ set() [2/3]

Nullable<T>& set ( const Nullable< T > & src )

inlineinherited

Set as copy of given value.

Parameters

src	Source to copy

Returns: This

◆ set() [3/3]

Nullable<T>& set ( T src )

inlineinherited

Set as given value.

Parameters

src	Source to set

Returns: This

◆ valid()

bool valid ( ) const

inlineinherited

Get whether valid (not null).

Returns: Whether valid

◆ value() [1/3]

const T& value ( ) const

inlineinherited

Get underlying value (const).

Alternative to operator*().

Returns: Whether valid

◆ value() [2/3]

T& value ( )

inlineinherited

Get underlying value.

Alternative to operator*().

Returns: Whether valid

◆ value() [3/3]

T value ( T defval ) const

inline

Get underlying value or given default if null.

Parameters

defval Default value to use if null

Returns: Result value, or defval if null

Member Data Documentation

◆ BYTES

const int BYTES = sizeof(T)

static

Type size in bytes.

◆ IS

const bool IS = IsFloat<T>::value

static

Whether type is really a floating point type.

◆ MAXDIGITS

const int MAXDIGITS = std::numeric_limits<T>::digits10

static

Maximum significant digits without precision loss.

◆ MAXDIGITS_AUTO

const int MAXDIGITS_AUTO = MAXDIGITS + 15

static

Max formatting digits with auto precision (used internally)

◆ NANOK

const bool NANOK = std::numeric_limits<T>::has_quiet_NaN

static

Whether Not-A-Number (NaN) is supported.

◆ SIGN

const bool SIGN = true

static

Whether type is signed.

The documentation for this struct was generated from the following file:

evo/type.h

Detailed Description

template<class T> struct evo::FloatT< T >

Public Types

Public Member Functions

Static Public Member Functions

Static Public Attributes

Member Typedef Documentation

◆ EvoNullableType

◆ This

◆ Type

Constructor & Destructor Documentation

◆ FloatT() [1/3]

◆ FloatT() [2/3]

◆ FloatT() [3/3]

Member Function Documentation

◆ compare() [1/2]

◆ compare() [2/2]

◆ denull()

◆ eps()

◆ eq() [1/2]

◆ eq() [2/2]

◆ eq1() [1/2]

◆ eq1() [2/2]

◆ exp10()

◆ fexp10()

◆ inf() [1/2]

◆ inf() [2/2]

◆ max()

◆ maxdigits_prec()

◆ maxexp()

◆ min()

◆ minexp()

◆ nan() [1/2]

◆ nan() [2/2]

◆ null()

◆ operator SafeBoolType()

◆ operator!()

◆ operator!=() [1/2]

◆ operator!=() [2/2]

◆ operator*() [1/2]

◆ operator*() [2/2]

◆ operator<() [1/2]

◆ operator<() [2/2]

◆ operator<=() [1/2]

◆ operator<=() [2/2]

◆ operator=() [1/3]

◆ operator=() [2/3]

◆ operator=() [3/3]

◆ operator==() [1/2]

◆ operator==() [2/2]

◆ operator>() [1/2]

◆ operator>() [2/2]

◆ operator>=() [1/2]

◆ operator>=() [2/2]

◆ precision()

◆ set() [1/3]

◆ set() [2/3]

◆ set() [3/3]

◆ valid()

◆ value() [1/3]

◆ value() [2/3]

◆ value() [3/3]

Member Data Documentation

◆ BYTES

◆ IS

◆ MAXDIGITS

◆ MAXDIGITS_AUTO

◆ NANOK

◆ SIGN

template<class T>
struct evo::FloatT< T >