Wrapped version of the C struct Pixel, with Alpha (opacity). More...

#include <i_pixel.h>

Inheritance diagram for kjb::PixelRGBA:

Public Member Functions
	PixelRGBA ()
	Default ctor, which leaves all fields uninitialized. More...

	PixelRGBA (const kjb_c::Pixel &p)
	This builds a valid pixel from another. More...

	PixelRGBA (float rr, float gg, float bb, float aa=255.0f)
	This builds a valid pixel with given values for color channels. More...

PixelRGBA &	ow_clamp ()
	Clamp all channels of the pixel to the range 0..255 inclusive. More...

PixelRGBA &	ow_clamp (float minval, float maxval)
	Clamp all channels to a specified range, overwriting. More...

PixelRGBA	clamp () const
	Return a new pixel with all channels clamped to range 0..255. More...

PixelRGBA	clamp (float minval, float maxval) const
	Return a new pixel with all channels clamped; non-overwriting. More...

PixelRGBA &	ow_floor ()
	Truncate pixel channel values to integers; overwriting. More...

PixelRGBA	floor () const
	Return new pixel with integer channel values. More...

PixelRGBA &	operator+= (const kjb_c::Pixel &p)
	Adds another pixel (channel-wise) to this pixel, overwriting. More...

PixelRGBA	operator+ (const kjb_c::Pixel &p) const
	This will add corresponding channels, creating a new pixel. More...

PixelRGBA &	operator*= (const PixelRGBA &p)
	Overwrite a pixel by multiplying it by p. More...

PixelRGBA &	operator*= (float k)
	Overwrite a pixel by scaling its value (r, g, b, and alpha too) More...

bool	operator== (const kjb_c::Pixel &p) const
	Test for EXACT pixel equality in all channels. More...

bool	operator!= (const kjb_c::Pixel &p) const
	Test for ANY pixel inequality in any channel. More...

std::string	as_hex_triplet () const
	express RGB values in an HTML-style string More...

Static Public Member Functions
static PixelRGBA	create_gray (float f)
	A "named constructor" to build a single opaque grayscale pixel. More...

static void	ow_clamp_channel (float *chan)
	Clamp one channel of the Pixel to the range 0..255 inclusive. More...

static void	ow_clamp_channel (float *chan, float minval, float maxval)
	Clamp one channel to a specified range. More...

Detailed Description

Wrapped version of the C struct Pixel, with Alpha (opacity).

This is a struct since there is (currently) no need to make such a simple item have anything private. The advantage of wrapping this at all is that there are commonly-performed operations that are naturally expressed as methods. Namely:

clamping
adding two pixels
scaling a pixel

Example: given pixels p, q this yields a Pixel: 0.5*(p.clamp()+q)

This struct assumes the user is not interested in the "Invalid Pixel" structure, and is perhaps (but not necessarily) using alpha (opacity). That's why the name has the RGBA suffix.

Remember that if you add pixels or scale pixels, the alpha channels also add and scale. So for the common case (opaque pixels, and 8-bit channels), the alpha channel is already at 255.0, so you might need to clamp() or to ow_clamp() after an add or scale operation, otherwise alpha could go overrange.

[CODE POLICE] TODO: Rename this class to follow kjb standard for type names (No camelcase Only first letter may be capitalized): Pixel_rgba

Constructor & Destructor Documentation

kjb::PixelRGBA::PixelRGBA ( )

inline

Default ctor, which leaves all fields uninitialized.

Warning: "Uninitialized" often means "bug-causing" and "troublesome."

kjb::PixelRGBA::PixelRGBA ( const kjb_c::Pixel & p )

inline

This builds a valid pixel from another.

Build a pixel from another. It receives a kjb_c::Pixel so as to work with PixelRGBA as well.

kjb::PixelRGBA::PixelRGBA	(	float	rr,
		float	gg,
		float	bb,
		float	aa = `255.0f`
	)

inline

This builds a valid pixel with given values for color channels.

Note that despite our massive dynamic range, the default value of aa (the alpha channel) is a mere 255.0f. This is the most common use-case, that of 8 bits of alpha, maximum opacity. However, it is ugly. For high dynamic range images, this is probably not your preferred value for the default opacity.

Member Function Documentation

std::string kjb::PixelRGBA::as_hex_triplet ( ) const

inline

express RGB values in an HTML-style string

PixelRGBA kjb::PixelRGBA::clamp ( ) const

inline

Return a new pixel with all channels clamped to range 0..255.

Observe that this routine does not, however, truncate pixels to integer values. If you need that, use the floor() method.

PixelRGBA kjb::PixelRGBA::clamp	(	float	minval,
		float	maxval
	)		const

inline

Return a new pixel with all channels clamped; non-overwriting.

static PixelRGBA kjb::PixelRGBA::create_gray ( float f )

inlinestatic

A "named constructor" to build a single opaque grayscale pixel.

PixelRGBA kjb::PixelRGBA::floor ( ) const

inline

Return new pixel with integer channel values.

bool kjb::PixelRGBA::operator!= ( const kjb_c::Pixel & p ) const

inline

Test for ANY pixel inequality in any channel.

Warning: Pixel stores channel values as floating point numbers! So instead of p != q you might want p.floor() != q.floor()

PixelRGBA& kjb::PixelRGBA::operator*= ( const PixelRGBA & p )

inline

Overwrite a pixel by multiplying it by p.

This performs a simple channel-wise product.

Warning: Don't forget to ow_clamp() if you think you might need to.

PixelRGBA& kjb::PixelRGBA::operator*= ( float k )

inline

Overwrite a pixel by scaling its value (r, g, b, and alpha too)

Warning: Don't forget to ow_clamp() if you think you might need to.

PixelRGBA kjb::PixelRGBA::operator+ ( const kjb_c::Pixel & p ) const

inline

This will add corresponding channels, creating a new pixel.

Warning: Don't forget to clamp() if you think you might need to.

PixelRGBA& kjb::PixelRGBA::operator+= ( const kjb_c::Pixel & p )

inline

Adds another pixel (channel-wise) to this pixel, overwriting.

Warning: Don't forget to ow_clamp() if you think you might need to.

bool kjb::PixelRGBA::operator== ( const kjb_c::Pixel & p ) const

inline

Test for EXACT pixel equality in all channels.

Warning: Pixel stores channel values as floating point numbers! So instead of p==q you might want p.floor()==q.floor()

PixelRGBA& kjb::PixelRGBA::ow_clamp ( )

inline

Clamp all channels of the pixel to the range 0..255 inclusive.

Returns: Lvalue of the modified PixelRGBA

Observe that this routine does not, however, truncate pixels to integer values. If you need that, use the ow_floor() or floor() method. This method has overwriting semantics.

PixelRGBA& kjb::PixelRGBA::ow_clamp	(	float	minval,
		float	maxval
	)

inline

Clamp all channels to a specified range, overwriting.

Returns: Lvalue of the modified PixelRGBA

static void kjb::PixelRGBA::ow_clamp_channel ( float * chan )

inlinestatic

Clamp one channel of the Pixel to the range 0..255 inclusive.

Observe that this routine does not, however, truncate pixels to integer values. If you need that, use the ow_floor() or floor() method. The 0..255 range is hardcoded in since it is such a common use case.

static void kjb::PixelRGBA::ow_clamp_channel	(	float *	chan,
		float	minval,
		float	maxval
	)

inlinestatic

Clamp one channel to a specified range.

PixelRGBA& kjb::PixelRGBA::ow_floor ( )

inline

Truncate pixel channel values to integers; overwriting.

Returns: Lvalue of the modified PixelRGBA

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

/work/kobus/src/lib/i_cpp/i_pixel.h

Public Member Functions

Static Public Member Functions

Detailed Description

Constructor & Destructor Documentation

Member Function Documentation