The fixed point converter provides conversion functions between a user type and up to 32 bit fixed point (signed or unsigned). More...

#include <FixedPointConverter.h>

Classes
struct	vectorToCooked_impl

struct	vectorToCooked_impl< std::string, RAW_ITERATOR, COOKED_ITERATOR >

Public Member Functions
	FixedPointConverter (std::string variableName, unsigned int nBits=32, int fractionalBits=0, bool isSignedFlag=true)
	The constructor defines the conversion factor. More...

template<typename UserType >
uint32_t	toRaw (UserType cookedValue) const
	Conversion function from type T to fixed point. More...

template<typename UserType , typename RAW_ITERATOR , typename COOKED_ITERATOR >
void	vectorToCooked (const RAW_ITERATOR &raw_begin, const RAW_ITERATOR &raw_end, const COOKED_ITERATOR &cooked_begin) const
	Conversion function from fixed-point values to type T. More...

template<typename UserType >
UserType	scalarToCooked (int32_t const &raw) const
	Inefficient convenience function for converting a single value to cooked. More...

unsigned int	getNBits () const
	Read back the number of bits the converter is using. More...

int	getFractionalBits () const
	Read back the fractional bits the converter is using. More...

bool	isSigned () const
	Read back wether the conversion is using signed values. More...

void	reconfigure (unsigned int nBits=32, int fractionalBits=0, bool isSignedFlag=true)
	Reconfigure the fixed point converter with new type information. More...

bool	operator== (const FixedPointConverter &other) const
	Compare two fixed point converters. More...

bool	operator!= (const FixedPointConverter &other) const

template<>
uint32_t	toRaw (Boolean cookedValue) const

template<>
uint32_t	toRaw (__attribute__((unused)) Void cookedValue) const

template<>
uint32_t	toRaw (std::string cookedValue) const

template<>
uint32_t	toRaw (Boolean cookedValue) const

template<>
uint32_t	toRaw (__attribute__((unused)) Void cookedValue) const

Detailed Description

The fixed point converter provides conversion functions between a user type and up to 32 bit fixed point (signed or unsigned).

Definition at line 28 of file FixedPointConverter.h.

Constructor & Destructor Documentation

◆ FixedPointConverter()

ChimeraTK::FixedPointConverter::FixedPointConverter	(	std::string	variableName,
		unsigned int	nBits = `32`,
		int	fractionalBits = `0`,
		bool	isSignedFlag = `true`
	)

explicit

The constructor defines the conversion factor.

Parameters

variableName	The name of the variable. It is used in case an exception is thrown which significantly simplifies the debugging.
nBits	The number of total bits must not exceed 32.
fractionalBits	The number of fractional bits can range from -1024+nBits to 1021-nBits (in case of a double user type). For integer user types, no fractional bits are allowed.
isSignedFlag	Flag whether the most significant bit is interpreted as sign bit. Negative numbers are interpreted as two's complement number of the respective number of bits (i.e. in signed 6 bit, 0 fractional bits 0x3F is -1)

Definition at line 14 of file FixedPointConverter.cc.

Here is the call graph for this function:

Member Function Documentation

◆ getFractionalBits()

int ChimeraTK::FixedPointConverter::getFractionalBits ( ) const

inline

Read back the fractional bits the converter is using.

Definition at line 96 of file FixedPointConverter.h.

Here is the caller graph for this function:

◆ getNBits()

unsigned int ChimeraTK::FixedPointConverter::getNBits ( ) const

inline

Read back the number of bits the converter is using.

Definition at line 93 of file FixedPointConverter.h.

Here is the caller graph for this function:

◆ isSigned()

bool ChimeraTK::FixedPointConverter::isSigned ( ) const

inline

Read back wether the conversion is using signed values.

Definition at line 99 of file FixedPointConverter.h.

Here is the caller graph for this function:

◆ operator!=()

bool ChimeraTK::FixedPointConverter::operator!= ( const FixedPointConverter & other ) const

inline

Definition at line 109 of file FixedPointConverter.h.

Here is the call graph for this function:

◆ operator==()

bool ChimeraTK::FixedPointConverter::operator== ( const FixedPointConverter & other ) const

inline

Compare two fixed point converters.

The variable name is ignored in this comparison.

Definition at line 106 of file FixedPointConverter.h.

Here is the caller graph for this function:

◆ reconfigure()

void ChimeraTK::FixedPointConverter::reconfigure	(	unsigned int	nBits = `32`,
		int	fractionalBits = `0`,
		bool	isSignedFlag = `true`
	)

Reconfigure the fixed point converter with new type information.

Definition at line 23 of file FixedPointConverter.cc.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ scalarToCooked()

template<typename UserType >

UserType ChimeraTK::FixedPointConverter::scalarToCooked ( int32_t const & raw ) const

inline

Inefficient convenience function for converting a single value to cooked.

Definition at line 86 of file FixedPointConverter.h.

Here is the caller graph for this function:

◆ toRaw() [1/6]

template<>

uint32_t ChimeraTK::FixedPointConverter::toRaw ( __attribute__((unused)) Void cookedValue ) const

Definition at line 110 of file FixedPointConverter.cc.

◆ toRaw() [2/6]

template<>

uint32_t ChimeraTK::FixedPointConverter::toRaw ( __attribute__((unused)) Void cookedValue ) const

◆ toRaw() [3/6]

template<>

uint32_t ChimeraTK::FixedPointConverter::toRaw ( Boolean cookedValue ) const

Definition at line 102 of file FixedPointConverter.cc.

◆ toRaw() [4/6]

template<>

uint32_t ChimeraTK::FixedPointConverter::toRaw ( Boolean cookedValue ) const

◆ toRaw() [5/6]

template<>

uint32_t ChimeraTK::FixedPointConverter::toRaw ( std::string cookedValue ) const

Definition at line 90 of file FixedPointConverter.cc.

◆ toRaw() [6/6]

template<typename UserType >

uint32_t ChimeraTK::FixedPointConverter::toRaw ( UserType cookedValue ) const

Conversion function from type T to fixed point.

This conversion usually will introduce rounding errors due to this limited resolution of the fixed point number compared to the double. In case of a high number of fractional bits this can mean that the most significant digits are cut and only the 'noise' in the less significant bits is represented.

Definition at line 336 of file FixedPointConverter.h.

Here is the caller graph for this function:

◆ vectorToCooked()

template<typename UserType , typename RAW_ITERATOR , typename COOKED_ITERATOR >

void ChimeraTK::FixedPointConverter::vectorToCooked	(	const RAW_ITERATOR &	raw_begin,
		const RAW_ITERATOR &	raw_end,
		const COOKED_ITERATOR &	cooked_begin
	)		const

inline

Conversion function from fixed-point values to type T.

The two vectors passed must be of equal size (i.e. cookedValues must already be properly allocated).

Definition at line 61 of file FixedPointConverter.h.

Here is the call graph for this function:

Here is the caller graph for this function:

The documentation for this class was generated from the following files:

/scratch/dragon/sources/ChimeraTK-DeviceAccess/include/FixedPointConverter.h
/scratch/dragon/sources/ChimeraTK-DeviceAccess/src/FixedPointConverter.cc

Classes

Public Member Functions

Detailed Description

Constructor & Destructor Documentation

◆ FixedPointConverter()

Member Function Documentation

◆ getFractionalBits()

◆ getNBits()

◆ isSigned()

◆ operator!=()

◆ operator==()

◆ reconfigure()

◆ scalarToCooked()

◆ toRaw() [1/6]

◆ toRaw() [2/6]

◆ toRaw() [3/6]

◆ toRaw() [4/6]

◆ toRaw() [5/6]

◆ toRaw() [6/6]

◆ vectorToCooked()