DCMTK Version 3.6.7
OFFIS DICOM Toolkit
Loading...
Searching...
No Matches
Static Public Member Functions | Static Public Attributes | List of all members
OFnumeric_limits< T > Struct Template Reference

A meta-template for querying various properties of fundamental types. More...

Static Public Member Functions

static T min ()
 The minimum finite value representable by the numeric type T.
 
static T lowest ()
 The lowest finite value representable by the numeric type T, that is, a finite value $x$ such that there is no other finite value $y$ where $y < x$.
 
static T max ()
 The maximum finite value representable by the numeric type T.
 
static T epsilon ()
 The machine epsilon, that is, the difference between $1.0$ and the next value representable by the floating-point type T.
 
static T round_error ()
 The largest possible rounding error in ULPs (units in the last place) as defined by ISO 10967, which can vary from $0.5$ (rounding to the nearest digit) to $1.0$ (rounding to zero or to infinity).
 
static T infinity ()
 The special value positive infinity, as represented by the floating-point type T.
 
static T quiet_NaN ()
 The special value quiet not-a-number, as represented by the floating-point type T.
 
static T signaling_NaN ()
 The special value signaling not-a-number, as represented by the floating-point type T.
 
static T denorm_min ()
 The minimum positive subnormal value of the type T, if OFnumeric_limits<T>::has_denorm != OFdenorm_absent, otherwise returns OFnumeric_limits<T>::min().
 

Static Public Attributes

static const OFBool is_specialized
 OFTrue for all T for which there exists a specialization of OFnumeric_limits, OFFalse otherwise.
 
static const OFBool is_signed
 OFTrue for all signed arithmetic types T and OFFalse for the unsigned types.
 
static const OFBool is_integer
 OFTrue for all integer arithmetic types T and OFFalse otherwise.
 
static const OFBool is_exact
 OFTrue for all arithmetic types T that use exact representation.
 
static const OFBool has_infinity
 OFTrue for all types T capable of representing the positive infinity as a distinct special value.
 
static const OFBool has_quiet_NaN
 OFTrue for all types T capable of representing the special value Quiet Not-A-Number.
 
static const OFBool has_signaling_NaN
 OFTrue for all types T capable of representing the special value Signaling Not-A-Number.
 
static const OFfloat_denorm_style has_denorm
 Identifies the floating-point types that support subnormal values.
 
static const OFBool has_denorm_loss
 OFTrue for all floating-point types T capable of distinguishing loss of precision due to denormalization from other causes of inexact result.
 
static const OFfloat_round_style round_style
 Identifies the rounding style used by the floating-point type T whenever a value that is not one of the exactly representable values of T is stored in an object of that type.
 
static const OFBool is_iec559
 OFTrue for all floating-point types T which fulfill the requirements of IEC 559 (IEEE 754) standard.
 
static const OFBool is_bounded
 OFTrue for all arithmetic types T that represent a finite set of values.
 
static const OFBool is_modulo
 OFTrue for all arithmetic types T that handle overflows with modulo arithmetic, that is, if the result of addition, subtraction, multiplication, or division of this type would fall outside the range $[min(), max()]$, the value returned by such operation differs from the expected value by a multiple of $max()-min() + 1$.
 
static const int digits
 The number of digits in base-radix that can be represented by the type T without change.
 
static const int digits10
 The number of base-10 digits that can be represented by the type T without change, that is, any number with this many decimal digits can be converted to a value of type T and back to decimal form, without change due to rounding or overflow.
 
static const int max_digits10
 The number of base-10 digits that are necessary to uniquely represent all distinct values of the type T, such as necessary for serialization/deserialization to text.
 
static const int radix
 The base of the number system used in the representation of the type.
 
static const int min_exponent
 The lowest negative number $n$ such that $radix^{n-1}$ is a valid normalized value of the floating-point type T.
 
static const int min_exponent10
 The lowest negative number $n$ such that $10^n$ is a valid normalized value of the floating-point type T.
 
static const int max_exponent
 The largest positive number $n$ such that $radix^{n-1}$ is a valid normalized value of the floating-point type T.
 
static const int max_exponent10
 The largest positive number $n$ such that $10^n$ is a valid normalized value of the floating-point type T.
 
static const OFBool traps
 OFTrue for all arithmetic types T that have at least one value that, if used as an argument to an arithmetic operation, will generate a trap.
 
static const OFBool tinyness_before
 OFTrue for all floating-point types T that test results of floating-point expressions for underflow before rounding.
 

Detailed Description

template<typename T>
struct OFnumeric_limits< T >

A meta-template for querying various properties of fundamental types.

The template OFnumeric_limits provides a standardized way to query various properties of fundamental types (e.g. the largest possible value for type int is OFnumeric_limits<int>::max()). OFnumeric_limits is compatible to C++11's std::numeric_limits.

Template Parameters
TThe type to inspect.

Member Function Documentation

◆ denorm_min()

template<typename T >
static T OFnumeric_limits< T >::denorm_min ( )
static

The minimum positive subnormal value of the type T, if OFnumeric_limits<T>::has_denorm != OFdenorm_absent, otherwise returns OFnumeric_limits<T>::min().

Note
Only meaningful for floating-point types.

C++11 standard definitions used:

T denorm_min()
non-specialized T()
OFBool OFFalse
char 0
signed char 0
unsigned char 0
signed short 0
unsigned short 0
signed int 0
unsigned int 0
signed long 0
unsigned long 0
float usually $2^{-149}$
double usually $2^{-1074}$

◆ epsilon()

template<typename T >
static T OFnumeric_limits< T >::epsilon ( )
static

The machine epsilon, that is, the difference between $1.0$ and the next value representable by the floating-point type T.

Note
It is only meaningful if OFnumeric_limits<T>::is_integer == OFFalse.

C++11 standard definitions used:

T epsilon()
non-specialized T()
OFBool OFFalse
char 0
signed char 0
unsigned char 0
signed short 0
unsigned short 0
signed int 0
unsigned int 0
signed long 0
unsigned long 0
float FLT_EPSILON
double DBL_EPSILON

◆ infinity()

template<typename T >
static T OFnumeric_limits< T >::infinity ( )
static

The special value positive infinity, as represented by the floating-point type T.

Note
Only meaningful if OFnumeric_limits<T>::has_infinity == OFTrue.
In IEEE 754, the most common binary representation of floating-point numbers, the positive infinity is the value with all bits of the exponent set and all bits of the fraction cleared.

C++11 standard definitions used:

T infinity()
non-specialized T()
OFBool OFFalse
char 0
signed char 0
unsigned char 0
signed short 0
unsigned short 0
signed int 0
unsigned int 0
signed long 0
unsigned long 0
float HUGE_VALF
double HUGE_VAL

◆ lowest()

template<typename T >
static T OFnumeric_limits< T >::lowest ( )
static

The lowest finite value representable by the numeric type T, that is, a finite value $x$ such that there is no other finite value $y$ where $y < x$.

Note
This is different from OFnumeric_limits<T>::min() for floating-point types.

C++11 standard definitions used:

T lowest()
non-specialized T()
OFBool OFFalse
char CHAR_MIN
signed char SCHAR_MIN
unsigned char 0
signed short SHRT_MIN
unsigned short 0
signed int INT_MIN
unsigned int 0
signed long LONG_MIN
unsigned long 0
float -FLT_MAX
double -DBL_MAX

◆ max()

template<typename T >
static T OFnumeric_limits< T >::max ( )
static

The maximum finite value representable by the numeric type T.

Note
Meaningful for all bounded types.

C++11 standard definitions used:

T max()
non-specialized T()
OFBool OFTrue
char CHAR_MAX
signed char SCHAR_MAX
unsigned char UCHAR_MAX
signed short SHRT_MAX
unsigned short USHRT_MAX
signed int INT_MAN
unsigned int UINT_MAX
signed long LONG_MAX
unsigned long ULONG_MAX
float FLT_MAX
double DBL_MAX

◆ min()

template<typename T >
static T OFnumeric_limits< T >::min ( )
static

The minimum finite value representable by the numeric type T.

For floating-point types with denormalization, min returns the minimum positive normalized value.

Note
This behavior may be unexpected, especially when compared to the behavior of min for integral types. To find the value that has no values less than it, use OFnumeric_limits<T>::lowest().
Min is only meaningful for bounded types and for unbounded unsigned types, that is, types that represent an infinite set of negative values have no meaningful minimum.

C++11 standard definitions used:

T min()
non-specialized T()
OFBool OFFalse
char CHAR_MIN
signed char SCHAR_MIN
unsigned char 0
signed short SHRT_MIN
unsigned short 0
signed int INT_MIN
unsigned int 0
signed long LONG_MIN
unsigned long 0
float FLT_MIN
double DBL_MIN

◆ quiet_NaN()

template<typename T >
static T OFnumeric_limits< T >::quiet_NaN ( )
static

The special value quiet not-a-number, as represented by the floating-point type T.

Note
Only meaningful if OFnumeric_limits<T>::has_quiet_NaN == OFTrue.
In IEEE 754, the most common binary representation of floating-point numbers, any value with all bits of the exponent set and at least one bit of the fraction set represents a NaN. It is implementation-defined which values of the fraction represent quiet or signaling NaNs, and whether the sign bit is meaningful.

C++11 standard definitions used:

T quiet_NaN()
non-specialized T()
OFBool OFFalse
char 0
signed char 0
unsigned char 0
signed short 0
unsigned short 0
signed int 0
unsigned int 0
signed long 0
unsigned long 0
float NAN or platform / compiler specific
double platform / compiler specific

◆ round_error()

template<typename T >
static T OFnumeric_limits< T >::round_error ( )
static

The largest possible rounding error in ULPs (units in the last place) as defined by ISO 10967, which can vary from $0.5$ (rounding to the nearest digit) to $1.0$ (rounding to zero or to infinity).

Note
It is only meaningful if OFnumeric_limits<T>::is_integer == OFFalse.

C++11 standard definitions used:

T round_error()
non-specialized T()
OFBool OFFalse
char 0
signed char 0
unsigned char 0
signed short 0
unsigned short 0
signed int 0
unsigned int 0
signed long 0
unsigned long 0
float 0.5f
double 0.5

◆ signaling_NaN()

template<typename T >
static T OFnumeric_limits< T >::signaling_NaN ( )
static

The special value signaling not-a-number, as represented by the floating-point type T.

Note
Only meaningful if OFnumeric_limits<T>::has_signaling_NaN == OFTrue.
In IEEE 754, the most common binary representation of floating-point numbers, any value with all bits of the exponent set and at least one bit of the fraction set represents a NaN. It is implementation-defined which values of the fraction represent quiet or signaling NaNs, and whether the sign bit is meaningful.

C++11 standard definitions used:

T signaling_NaN()
non-specialized T()
OFBool OFFalse
char 0
signed char 0
unsigned char 0
signed short 0
unsigned short 0
signed int 0
unsigned int 0
signed long 0
unsigned long 0
float platform / compiler specific
double platform / compiler specific

Member Data Documentation

◆ digits

template<typename T >
const int OFnumeric_limits< T >::digits
static

The number of digits in base-radix that can be represented by the type T without change.

For integer types, this is the number of bits not counting the sign bit. For floating-point types, this is the number of digits in the mantissa.

C++11 standard definitions used:

T digits
non-specialized 0
OFBool 1
char is_signed() ? CHAR_BIT : CHAR_BIT - 1
signed char CHAR_BIT - 1
unsigned char CHAR_BIT
signed short CHAR_BIT * sizeof(signed short) - 1
unsigned short CHAR_BIT * sizeof(unsigned short)
signed int CHAR_BIT * sizeof(signed int) - 1
unsigned int CHAR_BIT * sizeof(unsigned int)
signed long CHAR_BIT * sizeof(signed long) - 1
unsigned long CHAR_BIT * sizeof(unsigned long)
float FLT_MANT_DIG
double DBL_MANT_DIG

◆ digits10

template<typename T >
const int OFnumeric_limits< T >::digits10
static

The number of base-10 digits that can be represented by the type T without change, that is, any number with this many decimal digits can be converted to a value of type T and back to decimal form, without change due to rounding or overflow.

For base-radix types, it is the value of digits (digits-1 for floating-point types) multiplied by $log_{10}(radix)$ and rounded down.

C++11 standard definitions used:

T digits10
non-specialized 0
OFBool 0
char $digits \cdot log_{10}(2)$
signed char $digits \cdot log_{10}(2)$
unsigned char $digits \cdot log_{10}(2)$
signed short $digits \cdot log_{10}(2)$
unsigned short $digits \cdot log_{10}(2)$
signed int $digits \cdot log_{10}(2)$
unsigned int $digits \cdot log_{10}(2)$
signed long $digits \cdot log_{10}(2)$
unsigned long $digits \cdot log_{10}(2)$
float FLT_DIG
double DBL_DIG

◆ has_denorm

template<typename T >
const OFfloat_denorm_style OFnumeric_limits< T >::has_denorm
static

Identifies the floating-point types that support subnormal values.

C++11 standard definitions used:

T has_denorm
non-specialized OFdenorm_absent
OFBool OFdenorm_absent
char OFdenorm_absent
signed char OFdenorm_absent
unsigned char OFdenorm_absent
signed short OFdenorm_absent
unsigned short OFdenorm_absent
signed int OFdenorm_absent
unsigned int OFdenorm_absent
signed long OFdenorm_absent
unsigned long OFdenorm_absent
float usually OFdenorm_present
double usually OFdenorm_present

◆ has_denorm_loss

template<typename T >
const OFBool OFnumeric_limits< T >::has_denorm_loss
static

OFTrue for all floating-point types T capable of distinguishing loss of precision due to denormalization from other causes of inexact result.

Note
Standard-compliant IEEE 754 floating-point implementations may detect the floating-point underflow at three predefined moments:
  1. after computation of a result with absolute value smaller than OFnumeric_limits<T>::min(), such implementation detects tinyness before rounding.
  2. after rounding of the result to OFnumeric_limits<T>::digits bits, if the result is tiny, such implementation detects tinyness after rounding.
  3. if the conversion of the rounded tiny result to subnormal form resulted in the loss of precision, such implementation detects denorm loss.

C++11 standard definitions used:

T has_denorm_loss
non-specialized OFFalse
OFBool OFFalse
char OFFalse
signed char OFFalse
unsigned char OFFalse
signed short OFFalse
unsigned short OFFalse
signed int OFFalse
unsigned int OFFalse
signed long OFFalse
unsigned long OFFalse
float platform / compiler specific
double platform / compiler specific

◆ has_infinity

template<typename T >
const OFBool OFnumeric_limits< T >::has_infinity
static

OFTrue for all types T capable of representing the positive infinity as a distinct special value.

Note
This constant is meaningful for all floating-point types and is guaranteed to be OFTrue if OFnumeric_limits<T>::is_iec559 == OFTrue.

C++11 standard definitions used:

T has_infinity
non-specialized OFFalse
OFBool OFFalse
char OFFalse
signed char OFFalse
unsigned char OFFalse
signed short OFFalse
unsigned short OFFalse
signed int OFFalse
unsigned int OFFalse
signed long OFFalse
unsigned long OFFalse
float usually OFTrue
double usually OFTrue

◆ has_quiet_NaN

template<typename T >
const OFBool OFnumeric_limits< T >::has_quiet_NaN
static

OFTrue for all types T capable of representing the special value Quiet Not-A-Number.

Note
This constant is meaningful for all floating-point types and is guaranteed to be OFTrue if OFnumeric_limits<T>::is_iec559 == OFTrue.

C++11 standard definitions used:

T has_quit_NaN
non-specialized OFFalse
OFBool OFFalse
char OFFalse
signed char OFFalse
unsigned char OFFalse
signed short OFFalse
unsigned short OFFalse
signed int OFFalse
unsigned int OFFalse
signed long OFFalse
unsigned long OFFalse
float usually OFTrue
double usually OFTrue

◆ has_signaling_NaN

template<typename T >
const OFBool OFnumeric_limits< T >::has_signaling_NaN
static

OFTrue for all types T capable of representing the special value Signaling Not-A-Number.

Note
This constant is meaningful for all floating-point types and is guaranteed to be OFTrue if OFnumeric_limits<T>::is_iec559 == OFTrue.

C++11 standard definitions used:

T has_signaling_NaN
non-specialized OFFalse
OFBool OFFalse
char OFFalse
signed char OFFalse
unsigned char OFFalse
signed short OFFalse
unsigned short OFFalse
signed int OFFalse
unsigned int OFFalse
signed long OFFalse
unsigned long OFFalse
float usually OFTrue
double usually OFTrue

◆ is_bounded

template<typename T >
const OFBool OFnumeric_limits< T >::is_bounded
static

OFTrue for all arithmetic types T that represent a finite set of values.

Note
While all fundamental types are bounded, this constant would be OFFalse in a specialization of OFnumeric_limits for a library-provided arbitrary precision arithmetic type.

C++11 standard definitions used:

T is_bounded
non-specialized OFFalse
OFBool OFTrue
char OFTrue
signed char OFTrue
unsigned char OFTrue
signed short OFTrue
unsigned short OFTrue
signed int OFTrue
unsigned int OFTrue
signed long OFTrue
unsigned long OFTrue
float OFTrue
double OFTrue

◆ is_exact

template<typename T >
const OFBool OFnumeric_limits< T >::is_exact
static

OFTrue for all arithmetic types T that use exact representation.

Note
While all fundamental types T for which OFnumeric_limits<T>::is_exact == OFTrue are integer types, a library may define exact types that aren't integers, e.g. a rational arithmetics type representing fractions.

C++11 standard definitions used:

T is_exact
non-specialized OFFalse
OFBool OFTrue
char OFTrue
signed char OFTrue
unsigned char OFTrue
signed short OFTrue
unsigned short OFTrue
signed int OFTrue
unsigned int OFTrue
signed long OFTrue
unsigned long OFTrue
float OFFalse
double OFFalse

◆ is_iec559

template<typename T >
const OFBool OFnumeric_limits< T >::is_iec559
static

OFTrue for all floating-point types T which fulfill the requirements of IEC 559 (IEEE 754) standard.

Note
If OFnumeric_limits<T>::is_iec559 == OFTrue, then OFnumeric_limits<T>::has_infinity, OFnumeric_limits<T>::has_quiet_NaN, and OFnumeric_limits<T>::has_signaling_NaN are also OFTrue.

C++11 standard definitions used:

T is_iec559
non-specialized OFFalse
OFBool OFFalse
char OFFalse
signed char OFFalse
unsigned char OFFalse
signed short OFFalse
unsigned short OFFalse
signed int OFFalse
unsigned int OFFalse
signed long OFFalse
unsigned long OFFalse
float usually OFTrue
double usually OFTrue

◆ is_integer

template<typename T >
const OFBool OFnumeric_limits< T >::is_integer
static

OFTrue for all integer arithmetic types T and OFFalse otherwise.

C++11 standard definitions used:

T is_integer
non-specialized OFFalse
OFBool OFTrue
char OFTrue
signed char OFTrue
unsigned char OFTrue
signed short OFTrue
unsigned short OFTrue
signed int OFTrue
unsigned int OFTrue
signed long OFTrue
unsigned long OFTrue
float OFFalse
double OFFalse

◆ is_modulo

template<typename T >
const OFBool OFnumeric_limits< T >::is_modulo
static

OFTrue for all arithmetic types T that handle overflows with modulo arithmetic, that is, if the result of addition, subtraction, multiplication, or division of this type would fall outside the range $[min(), max()]$, the value returned by such operation differs from the expected value by a multiple of $max()-min() + 1$.

C++11 standard definitions used:

T is_modulo
non-specialized OFFalse
OFBool OFFalse
char platform / compiler specific
signed char platform / compiler specific
unsigned char OFTrue
signed short platform / compiler specific
unsigned short OFTrue
signed int platform / compiler specific
unsigned int OFTrue
signed long platform / compiler specific
unsigned long OFTrue
float OFFalse
double OFFalse

◆ is_signed

template<typename T >
const OFBool OFnumeric_limits< T >::is_signed
static

OFTrue for all signed arithmetic types T and OFFalse for the unsigned types.

C++11 standard definitions used:

T is_signed
non-specialized OFFalse
OFBool OFFalse
char platform / compiler specific
signed char OFTrue
unsigned char OFFalse
signed short OFTrue
unsigned short OFFalse
signed int OFTrue
unsigned int OFFalse
signed long OFTrue
unsigned long OFFalse
float OFTrue
double OFTrue

◆ is_specialized

template<typename T >
const OFBool OFnumeric_limits< T >::is_specialized
static

OFTrue for all T for which there exists a specialization of OFnumeric_limits, OFFalse otherwise.

C++11 standard definitions used:

T is_specialized
non-specialized OFFalse
OFBool OFTrue
char OFTrue
signed char OFTrue
unsigned char OFTrue
signed short OFTrue
unsigned short OFTrue
signed int OFTrue
unsigned int OFTrue
signed long OFTrue
unsigned long OFTrue
float OFTrue
double OFTrue

◆ max_digits10

template<typename T >
const int OFnumeric_limits< T >::max_digits10
static

The number of base-10 digits that are necessary to uniquely represent all distinct values of the type T, such as necessary for serialization/deserialization to text.

Note
This constant is meaningful for all floating-point types.

C++11 standard definitions used:

T max_digits10
non-specialized 0
OFBool 0
char 0
signed char 0
unsigned char 0
signed short 0
unsigned short 0
signed int 0
unsigned int 0
signed long 0
unsigned long 0
float $\lfloor digits \cdot log_{10}(2) + 2 \rfloor$
double $\lfloor digits \cdot log_{10}(2) + 2 \rfloor$

◆ max_exponent

template<typename T >
const int OFnumeric_limits< T >::max_exponent
static

The largest positive number $n$ such that $radix^{n-1}$ is a valid normalized value of the floating-point type T.

C++11 standard definitions used:

T max_exponent
non-specialized 0
OFBool 0
char 0
signed char 0
unsigned char 0
signed short 0
unsigned short 0
signed int 0
unsigned int 0
signed long 0
unsigned long 0
float FLT_MAX_EXP
double DBL_MAX_EXP

◆ max_exponent10

template<typename T >
const int OFnumeric_limits< T >::max_exponent10
static

The largest positive number $n$ such that $10^n$ is a valid normalized value of the floating-point type T.

C++11 standard definitions used:

T max_exponent10
non-specialized 0
OFBool 0
char 0
signed char 0
unsigned char 0
signed short 0
unsigned short 0
signed int 0
unsigned int 0
signed long 0
unsigned long 0
float FLT_MAX_10_EXP
double DBL_MAX_10_EXP

◆ min_exponent

template<typename T >
const int OFnumeric_limits< T >::min_exponent
static

The lowest negative number $n$ such that $radix^{n-1}$ is a valid normalized value of the floating-point type T.

C++11 standard definitions used:

T min_exponent
non-specialized 0
OFBool 0
char 0
signed char 0
unsigned char 0
signed short 0
unsigned short 0
signed int 0
unsigned int 0
signed long 0
unsigned long 0
float FLT_MIN_EXP
double DBL_MIN_EXP

◆ min_exponent10

template<typename T >
const int OFnumeric_limits< T >::min_exponent10
static

The lowest negative number $n$ such that $10^n$ is a valid normalized value of the floating-point type T.

C++11 standard definitions used:

T min_exponent10
non-specialized 0
OFBool 0
char 0
signed char 0
unsigned char 0
signed short 0
unsigned short 0
signed int 0
unsigned int 0
signed long 0
unsigned long 0
float FLT_MIN_10_EXP
double DBL_MIN_10_EXP

◆ radix

template<typename T >
const int OFnumeric_limits< T >::radix
static

The base of the number system used in the representation of the type.

It is 2 for all binary numeric types, but it may be, for example, 10 for IEEE 754 decimal floating-point types or for third-party binary-coded decimal integers. This constant is meaningful for all specializations.

C++11 standard definitions used:

T radix
non-specialized 0
OFBool 2
char 2
signed char 2
unsigned char 2
signed short 2
unsigned short 2
signed int 2
unsigned int 2
signed long 2
unsigned long 2
float FLT_RADIX
double FLT_RADIX

◆ round_style

template<typename T >
const OFfloat_round_style OFnumeric_limits< T >::round_style
static

Identifies the rounding style used by the floating-point type T whenever a value that is not one of the exactly representable values of T is stored in an object of that type.

C++11 standard definitions used:

T round_style
non-specialized OFround_toward_zero
OFBool OFround_toward_zero
char OFround_toward_zero
signed char OFround_toward_zero
unsigned char OFround_toward_zero
signed short OFround_toward_zero
unsigned short OFround_toward_zero
signed int OFround_toward_zero
unsigned int OFround_toward_zero
signed long OFround_toward_zero
unsigned long OFround_toward_zero
float usually OFround_to_nearest
double usually OFround_to_nearest

◆ tinyness_before

template<typename T >
const OFBool OFnumeric_limits< T >::tinyness_before
static

OFTrue for all floating-point types T that test results of floating-point expressions for underflow before rounding.

Note
Standard-compliant IEEE 754 floating-point implementations may detect the floating-point underflow at three predefined moments:
  1. after computation of a result with absolute value smaller than OFnumeric_limits<T>::min(), such implementation detects tinyness before rounding.
  2. after rounding of the result to OFnumeric_limits<T>::digits bits, if the result is tiny, such implementation detects tinyness after rounding.
  3. if the conversion of the rounded tiny result to subnormal form resulted in the loss of precision, such implementation detects denorm loss.

C++11 standard definitions used:

T tinyness_before
non-specialized OFFalse
OFBool OFFalse
char OFFalse
signed char OFFalse
unsigned char OFFalse
signed short OFFalse
unsigned short OFFalse
signed int OFFalse
unsigned int OFFalse
signed long OFFalse
unsigned long OFFalse
float platform / compiler specific
double platform / compiler specific

◆ traps

template<typename T >
const OFBool OFnumeric_limits< T >::traps
static

OFTrue for all arithmetic types T that have at least one value that, if used as an argument to an arithmetic operation, will generate a trap.

Note
On most platforms integer division by zero always traps, and OFnumeric_limits<T>::traps is OFTrue for all integer types that support the value 0. The exception is the type OFBool, if support for the native type bool is available: even though division by OFFalse traps due to integral promotion from OFBool to int, it is the zero-valued int that traps. Zero is not a value of type bool.
On most platforms, floating-point exceptions may be turned on and off at runtime, in which case the value of OFnumeric_limits<T>::traps for floating-point types reflects the state of floating-point trapping facility at the time of program startup.

C++11 standard definitions used:

T traps
non-specialized OFFalse
OFBool usually OFFalse
char usually OFTrue
signed char usually OFTrue
unsigned char usually OFTrue
signed short usually OFTrue
unsigned short usually OFTrue
signed int usually OFTrue
unsigned int usually OFTrue
signed long usually OFTrue
unsigned long usually OFTrue
float usually OFFalse
double usually OFFalse
The documentation for this struct was generated from the following file:

Generated on Mon Apr 15 2024 for DCMTK Version 3.6.7 by Doxygen 1.9.8