ic
/
magnum
mirror of https://github.com/mosra/magnum.git
6
0
Fork 0
Code Issues Projects Releases Wiki Activity
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
10487 Commits
21 Branches
19 Tags
70 MiB
Tree: 21cff30b27
Branches Tags
${ item.name }
Create tag ${ searchTerm }
Create branch ${ searchTerm }
from '21cff30b27'
${ noResults }
magnum/src/Magnum/PixelFormat.cpp

873 lines
32 KiB
Raw Normal View History Unescape

Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
/*
This file is part of Magnum.
Updated copyright year.
6 years ago
Copyright © 2010, 2011, 2012, 2013, 2014, 2015, 2016, 2017, 2018, 2019,
Updated copyright year.
4 years ago
2020, 2021, 2022 Vladimír Vondruš <mosra@centrum.cz>
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
Permission is hereby granted, free of charge, to any person obtaining a
copy of this software and associated documentation files (the "Software"),
to deal in the Software without restriction, including without limitation
the rights to use, copy, modify, merge, publish, distribute, sublicense,
and/or sell copies of the Software, and to permit persons to whom the
Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included
in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
DEALINGS IN THE SOFTWARE.
*/
Minor cleanup.
7 years ago
#include "PixelFormat.h"
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
Use Corrade String[View] in some ConfigurationValue specializations. These headers are included quite often and because <Corrade/Utility/StlForwardString.h> expands to the whole <string>, it has quite a significant impact. Other cases like Math/ConfigurationValue.h are not changed at the moment, as that header is only included together with Utility::Arguments or Utility::Configuration -- and these classes are still STL-heavy, so avoiding std::string usage in just the ConfigurationValue specializations doesn't help anything at the moment.
4 years ago
#include <Corrade/Containers/String.h>
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
#include <Corrade/Containers/ArrayView.h>
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
#include <Corrade/Utility/Assert.h>
#include <Corrade/Utility/Debug.h>
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
#include "Magnum/Math/Vector3.h"
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
namespace Magnum {
Rename pixelSize() and other helpers to pixelFormatSize() etc. For consistency with how VertexFormat and other enum helpers are named. The compressedBlockSize() and compressedBlockDataSize() is also renamed to compressedPixelFormatBlockSize() and compressedPixelFormatBlockDataSize(). While backwards compatibility aliases are in place, a breaking change is that Image classes now look for pixelFormatSize() instead of pixelSize(). This is used e.g. when passing GL::PixelFormat / GL::PixelType to the image classes, instead of the generic PixelFormat. While useful, it's unlikely that any project was defining their own pixel format enum and pixelSize() for a D3D or Metal renderer or whatnot, so the breakage should have no practical impact.
4 years ago
UnsignedInt pixelFormatSize(const PixelFormat format) {
Make various PixelFormat assertions print the actual value. Otherwise they are pretty useless.
6 years ago
CORRADE_ASSERT(!isPixelFormatImplementationSpecific(format),
Rename pixelSize() and other helpers to pixelFormatSize() etc. For consistency with how VertexFormat and other enum helpers are named. The compressedBlockSize() and compressedBlockDataSize() is also renamed to compressedPixelFormatBlockSize() and compressedPixelFormatBlockDataSize(). While backwards compatibility aliases are in place, a breaking change is that Image classes now look for pixelFormatSize() instead of pixelSize(). This is used e.g. when passing GL::PixelFormat / GL::PixelType to the image classes, instead of the generic PixelFormat. While useful, it's unlikely that any project was defining their own pixel format enum and pixelSize() for a D3D or Metal renderer or whatnot, so the breakage should have no practical impact.
4 years ago
"pixelFormatSize(): can't determine size of an implementation-specific format" << reinterpret_cast<void*>(pixelFormatUnwrap(format)), {});
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
Prefer CORRADE_TARGET_* instead of __GNUC__, __clang__ and _MSVC_VER. While branching on a compiler is rather common, checking a particular compiler version should be needed only rarely. Thus minimize use of such macros to make them easier to grep for.
4 years ago
#ifdef CORRADE_TARGET_GCC
Be strict for missing values in pixel format switches.
7 years ago
#pragma GCC diagnostic push
#pragma GCC diagnostic error "-Wswitch"
#endif
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
switch(format) {
case PixelFormat::R8Unorm:
case PixelFormat::R8Snorm:
Added sRGB formats to PixelFormats + R and RG sRGB GL texture formats. The extension support is a mess, as usual. This time the ES flavor supports something that desktop doesn't and WebGL lack those altogether. Yay.
7 years ago
case PixelFormat::R8Srgb:
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
case PixelFormat::R8UI:
case PixelFormat::R8I:
Add generic depth/stencil PixelFormat values. The 32-bit float depth be needed for the upcoming OpenEXR plugin, added also the remaining ones that will be eventually supported by KTX and DDS plugins.
5 years ago
case PixelFormat::Stencil8UI:
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
return 1;
case PixelFormat::RG8Unorm:
case PixelFormat::RG8Snorm:
Added sRGB formats to PixelFormats + R and RG sRGB GL texture formats. The extension support is a mess, as usual. This time the ES flavor supports something that desktop doesn't and WebGL lack those altogether. Yay.
7 years ago
case PixelFormat::RG8Srgb:
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
case PixelFormat::RG8UI:
case PixelFormat::RG8I:
case PixelFormat::R16Unorm:
case PixelFormat::R16Snorm:
case PixelFormat::R16UI:
case PixelFormat::R16I:
case PixelFormat::R16F:
Add generic depth/stencil PixelFormat values. The 32-bit float depth be needed for the upcoming OpenEXR plugin, added also the remaining ones that will be eventually supported by KTX and DDS plugins.
5 years ago
case PixelFormat::Depth16Unorm:
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
return 2;
case PixelFormat::RGB8Unorm:
case PixelFormat::RGB8Snorm:
Added sRGB formats to PixelFormats + R and RG sRGB GL texture formats. The extension support is a mess, as usual. This time the ES flavor supports something that desktop doesn't and WebGL lack those altogether. Yay.
7 years ago
case PixelFormat::RGB8Srgb:
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
case PixelFormat::RGB8UI:
case PixelFormat::RGB8I:
return 3;
case PixelFormat::RGBA8Unorm:
case PixelFormat::RGBA8Snorm:
Added sRGB formats to PixelFormats + R and RG sRGB GL texture formats. The extension support is a mess, as usual. This time the ES flavor supports something that desktop doesn't and WebGL lack those altogether. Yay.
7 years ago
case PixelFormat::RGBA8Srgb:
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
case PixelFormat::RGBA8UI:
case PixelFormat::RGBA8I:
case PixelFormat::RG16Unorm:
case PixelFormat::RG16Snorm:
case PixelFormat::RG16UI:
case PixelFormat::RG16I:
case PixelFormat::RG16F:
case PixelFormat::R32UI:
case PixelFormat::R32I:
case PixelFormat::R32F:
Add generic depth/stencil PixelFormat values. The 32-bit float depth be needed for the upcoming OpenEXR plugin, added also the remaining ones that will be eventually supported by KTX and DDS plugins.
5 years ago
case PixelFormat::Depth24Unorm:
case PixelFormat::Depth32F:
case PixelFormat::Depth16UnormStencil8UI:
case PixelFormat::Depth24UnormStencil8UI:
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
return 4;
case PixelFormat::RGB16Unorm:
case PixelFormat::RGB16Snorm:
case PixelFormat::RGB16UI:
case PixelFormat::RGB16I:
case PixelFormat::RGB16F:
return 6;
case PixelFormat::RGBA16Unorm:
case PixelFormat::RGBA16Snorm:
case PixelFormat::RGBA16UI:
case PixelFormat::RGBA16I:
case PixelFormat::RGBA16F:
case PixelFormat::RG32UI:
case PixelFormat::RG32I:
case PixelFormat::RG32F:
Add generic depth/stencil PixelFormat values. The 32-bit float depth be needed for the upcoming OpenEXR plugin, added also the remaining ones that will be eventually supported by KTX and DDS plugins.
5 years ago
case PixelFormat::Depth32FStencil8UI:
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
return 8;
case PixelFormat::RGB32UI:
case PixelFormat::RGB32I:
case PixelFormat::RGB32F:
return 12;
case PixelFormat::RGBA32UI:
case PixelFormat::RGBA32I:
case PixelFormat::RGBA32F:
return 16;
}
Prefer CORRADE_TARGET_* instead of __GNUC__, __clang__ and _MSVC_VER. While branching on a compiler is rather common, checking a particular compiler version should be needed only rarely. Thus minimize use of such macros to make them easier to grep for.
4 years ago
#ifdef CORRADE_TARGET_GCC
Be strict for missing values in pixel format switches.
7 years ago
#pragma GCC diagnostic pop
#endif
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
Rename pixelSize() and other helpers to pixelFormatSize() etc. For consistency with how VertexFormat and other enum helpers are named. The compressedBlockSize() and compressedBlockDataSize() is also renamed to compressedPixelFormatBlockSize() and compressedPixelFormatBlockDataSize(). While backwards compatibility aliases are in place, a breaking change is that Image classes now look for pixelFormatSize() instead of pixelSize(). This is used e.g. when passing GL::PixelFormat / GL::PixelType to the image classes, instead of the generic PixelFormat. While useful, it's unlikely that any project was defining their own pixel format enum and pixelSize() for a D3D or Metal renderer or whatnot, so the breakage should have no practical impact.
4 years ago
CORRADE_ASSERT_UNREACHABLE("pixelFormatSize(): invalid format" << format, {});
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
}
Add PixelFormat decomposition utilities. Similar to what VertexFormat already has -- getting channel count, channel format, sRGB and color vs depth/stencil properties out of a PixelFormat value. Very useful for various image conversion plugins.
4 years ago
PixelFormat pixelFormatChannelFormat(const PixelFormat format) {
CORRADE_ASSERT(!isPixelFormatImplementationSpecific(format),
"pixelFormatChannelFormat(): can't determine channel format of an implementation-specific format" << reinterpret_cast<void*>(pixelFormatUnwrap(format)), {});
#ifdef CORRADE_TARGET_GCC
#pragma GCC diagnostic push
#pragma GCC diagnostic error "-Wswitch"
#endif
switch(format) {
case PixelFormat::R8Unorm:
case PixelFormat::RG8Unorm:
case PixelFormat::RGB8Unorm:
case PixelFormat::RGBA8Unorm:
return PixelFormat::R8Unorm;
case PixelFormat::R8Snorm:
case PixelFormat::RG8Snorm:
case PixelFormat::RGB8Snorm:
case PixelFormat::RGBA8Snorm:
return PixelFormat::R8Snorm;
case PixelFormat::R8Srgb:
case PixelFormat::RG8Srgb:
case PixelFormat::RGB8Srgb:
case PixelFormat::RGBA8Srgb:
return PixelFormat::R8Srgb;
case PixelFormat::R8UI:
case PixelFormat::RG8UI:
case PixelFormat::RGB8UI:
case PixelFormat::RGBA8UI:
return PixelFormat::R8UI;
case PixelFormat::R8I:
case PixelFormat::RG8I:
case PixelFormat::RGB8I:
case PixelFormat::RGBA8I:
return PixelFormat::R8I;
case PixelFormat::R16Unorm:
case PixelFormat::RG16Unorm:
case PixelFormat::RGB16Unorm:
case PixelFormat::RGBA16Unorm:
return PixelFormat::R16Unorm;
case PixelFormat::R16Snorm:
case PixelFormat::RG16Snorm:
case PixelFormat::RGB16Snorm:
case PixelFormat::RGBA16Snorm:
return PixelFormat::R16Snorm;
case PixelFormat::R16UI:
case PixelFormat::RG16UI:
case PixelFormat::RGB16UI:
case PixelFormat::RGBA16UI:
return PixelFormat::R16UI;
case PixelFormat::R16I:
case PixelFormat::RG16I:
case PixelFormat::RGB16I:
case PixelFormat::RGBA16I:
return PixelFormat::R16I;
case PixelFormat::R32UI:
case PixelFormat::RG32UI:
case PixelFormat::RGB32UI:
case PixelFormat::RGBA32UI:
return PixelFormat::R32UI;
case PixelFormat::R32I:
case PixelFormat::RG32I:
case PixelFormat::RGB32I:
case PixelFormat::RGBA32I:
return PixelFormat::R32I;
case PixelFormat::R16F:
case PixelFormat::RG16F:
case PixelFormat::RGB16F:
case PixelFormat::RGBA16F:
return PixelFormat::R16F;
case PixelFormat::R32F:
case PixelFormat::RG32F:
case PixelFormat::RGB32F:
case PixelFormat::RGBA32F:
return PixelFormat::R32F;
case PixelFormat::Stencil8UI:
case PixelFormat::Depth16Unorm:
case PixelFormat::Depth24Unorm:
case PixelFormat::Depth32F:
case PixelFormat::Depth16UnormStencil8UI:
case PixelFormat::Depth24UnormStencil8UI:
case PixelFormat::Depth32FStencil8UI:
CORRADE_ASSERT_UNREACHABLE("pixelFormatChannelFormat(): can't determine channel format of" << format, {});
}
#ifdef CORRADE_TARGET_GCC
#pragma GCC diagnostic pop
#endif
CORRADE_ASSERT_UNREACHABLE("pixelFormatChannelFormat(): invalid format" << format, {});
}
UnsignedInt pixelFormatChannelCount(const PixelFormat format) {
CORRADE_ASSERT(!isPixelFormatImplementationSpecific(format),
"pixelFormatChannelCount(): can't determine channel count of an implementation-specific format" << reinterpret_cast<void*>(pixelFormatUnwrap(format)), {});
#ifdef CORRADE_TARGET_GCC
#pragma GCC diagnostic push
#pragma GCC diagnostic error "-Wswitch"
#endif
switch(format) {
case PixelFormat::R8Unorm:
case PixelFormat::R8Snorm:
case PixelFormat::R8Srgb:
case PixelFormat::R8UI:
case PixelFormat::R8I:
case PixelFormat::R16Unorm:
case PixelFormat::R16Snorm:
case PixelFormat::R16UI:
case PixelFormat::R16I:
case PixelFormat::R32UI:
case PixelFormat::R32I:
case PixelFormat::R16F:
case PixelFormat::R32F:
return 1;
case PixelFormat::RG8Unorm:
case PixelFormat::RG8Snorm:
case PixelFormat::RG8Srgb:
case PixelFormat::RG8UI:
case PixelFormat::RG8I:
case PixelFormat::RG16Unorm:
case PixelFormat::RG16Snorm:
case PixelFormat::RG16UI:
case PixelFormat::RG16I:
case PixelFormat::RG32UI:
case PixelFormat::RG32I:
case PixelFormat::RG16F:
case PixelFormat::RG32F:
return 2;
case PixelFormat::RGB8Unorm:
case PixelFormat::RGB8Snorm:
case PixelFormat::RGB8Srgb:
case PixelFormat::RGB8UI:
case PixelFormat::RGB8I:
case PixelFormat::RGB16Unorm:
case PixelFormat::RGB16Snorm:
case PixelFormat::RGB16UI:
case PixelFormat::RGB16I:
case PixelFormat::RGB32UI:
case PixelFormat::RGB32I:
case PixelFormat::RGB16F:
case PixelFormat::RGB32F:
return 3;
case PixelFormat::RGBA8Unorm:
case PixelFormat::RGBA8Snorm:
case PixelFormat::RGBA8Srgb:
case PixelFormat::RGBA8UI:
case PixelFormat::RGBA8I:
case PixelFormat::RGBA16Unorm:
case PixelFormat::RGBA16Snorm:
case PixelFormat::RGBA16UI:
case PixelFormat::RGBA16I:
case PixelFormat::RGBA32UI:
case PixelFormat::RGBA32I:
case PixelFormat::RGBA16F:
case PixelFormat::RGBA32F:
return 4;
case PixelFormat::Depth16Unorm:
case PixelFormat::Depth24Unorm:
case PixelFormat::Depth32F:
case PixelFormat::Stencil8UI:
case PixelFormat::Depth16UnormStencil8UI:
case PixelFormat::Depth24UnormStencil8UI:
case PixelFormat::Depth32FStencil8UI:
CORRADE_ASSERT_UNREACHABLE("pixelFormatChannelCount(): can't determine channel count of" << format, {});
}
#ifdef CORRADE_TARGET_GCC
#pragma GCC diagnostic pop
#endif
CORRADE_ASSERT_UNREACHABLE("pixelFormatChannelCount(): invalid format" << format, {});
}
Add isPixelFormat{Normalized,Integer,FloatingPoint}() helpers.
3 years ago
bool isPixelFormatNormalized(const PixelFormat format) {
CORRADE_ASSERT(!isPixelFormatImplementationSpecific(format),
"isPixelFormatNormalized(): can't determine type of an implementation-specific format" << reinterpret_cast<void*>(pixelFormatUnwrap(format)), {});
#ifdef CORRADE_TARGET_GCC
#pragma GCC diagnostic push
#pragma GCC diagnostic error "-Wswitch"
#endif
switch(format) {
case PixelFormat::R8Unorm:
case PixelFormat::RG8Unorm:
case PixelFormat::RGB8Unorm:
case PixelFormat::RGBA8Unorm:
case PixelFormat::R8Snorm:
case PixelFormat::RG8Snorm:
case PixelFormat::RGB8Snorm:
case PixelFormat::RGBA8Snorm:
case PixelFormat::R8Srgb:
case PixelFormat::RG8Srgb:
case PixelFormat::RGB8Srgb:
case PixelFormat::RGBA8Srgb:
case PixelFormat::R16Unorm:
case PixelFormat::RG16Unorm:
case PixelFormat::RGB16Unorm:
case PixelFormat::RGBA16Unorm:
case PixelFormat::R16Snorm:
case PixelFormat::RG16Snorm:
case PixelFormat::RGB16Snorm:
case PixelFormat::RGBA16Snorm:
return true;
case PixelFormat::R8UI:
case PixelFormat::RG8UI:
case PixelFormat::RGB8UI:
case PixelFormat::RGBA8UI:
case PixelFormat::R8I:
case PixelFormat::RG8I:
case PixelFormat::RGB8I:
case PixelFormat::RGBA8I:
case PixelFormat::R16UI:
case PixelFormat::RG16UI:
case PixelFormat::RGB16UI:
case PixelFormat::RGBA16UI:
case PixelFormat::R16I:
case PixelFormat::RG16I:
case PixelFormat::RGB16I:
case PixelFormat::RGBA16I:
case PixelFormat::R32UI:
case PixelFormat::RG32UI:
case PixelFormat::RGB32UI:
case PixelFormat::RGBA32UI:
case PixelFormat::R32I:
case PixelFormat::RG32I:
case PixelFormat::RGB32I:
case PixelFormat::RGBA32I:
case PixelFormat::R16F:
case PixelFormat::RG16F:
case PixelFormat::RGB16F:
case PixelFormat::RGBA16F:
case PixelFormat::R32F:
case PixelFormat::RG32F:
case PixelFormat::RGB32F:
case PixelFormat::RGBA32F:
return false;
case PixelFormat::Depth16Unorm:
case PixelFormat::Depth24Unorm:
case PixelFormat::Depth32F:
case PixelFormat::Stencil8UI:
case PixelFormat::Depth16UnormStencil8UI:
case PixelFormat::Depth24UnormStencil8UI:
case PixelFormat::Depth32FStencil8UI:
CORRADE_ASSERT_UNREACHABLE("isPixelFormatNormalized(): can't determine type of" << format, {});
}
#ifdef CORRADE_TARGET_GCC
#pragma GCC diagnostic pop
#endif
CORRADE_ASSERT_UNREACHABLE("isPixelFormatNormalized(): invalid format" << format, {});
}
bool isPixelFormatIntegral(const PixelFormat format) {
CORRADE_ASSERT(!isPixelFormatImplementationSpecific(format),
"isPixelFormatIntegral(): can't determine type of an implementation-specific format" << reinterpret_cast<void*>(pixelFormatUnwrap(format)), {});
#ifdef CORRADE_TARGET_GCC
#pragma GCC diagnostic push
#pragma GCC diagnostic error "-Wswitch"
#endif
switch(format) {
case PixelFormat::R8UI:
case PixelFormat::RG8UI:
case PixelFormat::RGB8UI:
case PixelFormat::RGBA8UI:
case PixelFormat::R8I:
case PixelFormat::RG8I:
case PixelFormat::RGB8I:
case PixelFormat::RGBA8I:
case PixelFormat::R16UI:
case PixelFormat::RG16UI:
case PixelFormat::RGB16UI:
case PixelFormat::RGBA16UI:
case PixelFormat::R16I:
case PixelFormat::RG16I:
case PixelFormat::RGB16I:
case PixelFormat::RGBA16I:
case PixelFormat::R32UI:
case PixelFormat::RG32UI:
case PixelFormat::RGB32UI:
case PixelFormat::RGBA32UI:
case PixelFormat::R32I:
case PixelFormat::RG32I:
case PixelFormat::RGB32I:
case PixelFormat::RGBA32I:
return true;
case PixelFormat::R8Unorm:
case PixelFormat::RG8Unorm:
case PixelFormat::RGB8Unorm:
case PixelFormat::RGBA8Unorm:
case PixelFormat::R8Snorm:
case PixelFormat::RG8Snorm:
case PixelFormat::RGB8Snorm:
case PixelFormat::RGBA8Snorm:
case PixelFormat::R8Srgb:
case PixelFormat::RG8Srgb:
case PixelFormat::RGB8Srgb:
case PixelFormat::RGBA8Srgb:
case PixelFormat::R16Unorm:
case PixelFormat::RG16Unorm:
case PixelFormat::RGB16Unorm:
case PixelFormat::RGBA16Unorm:
case PixelFormat::R16Snorm:
case PixelFormat::RG16Snorm:
case PixelFormat::RGB16Snorm:
case PixelFormat::RGBA16Snorm:
case PixelFormat::R16F:
case PixelFormat::RG16F:
case PixelFormat::RGB16F:
case PixelFormat::RGBA16F:
case PixelFormat::R32F:
case PixelFormat::RG32F:
case PixelFormat::RGB32F:
case PixelFormat::RGBA32F:
return false;
case PixelFormat::Depth16Unorm:
case PixelFormat::Depth24Unorm:
case PixelFormat::Depth32F:
case PixelFormat::Stencil8UI:
case PixelFormat::Depth16UnormStencil8UI:
case PixelFormat::Depth24UnormStencil8UI:
case PixelFormat::Depth32FStencil8UI:
CORRADE_ASSERT_UNREACHABLE("isPixelFormatIntegral(): can't determine type of" << format, {});
}
#ifdef CORRADE_TARGET_GCC
#pragma GCC diagnostic pop
#endif
CORRADE_ASSERT_UNREACHABLE("isPixelFormatIntegral(): invalid format" << format, {});
}
bool isPixelFormatFloatingPoint(const PixelFormat format) {
CORRADE_ASSERT(!isPixelFormatImplementationSpecific(format),
"isPixelFormatFloatingPoint(): can't determine type of an implementation-specific format" << reinterpret_cast<void*>(pixelFormatUnwrap(format)), {});
#ifdef CORRADE_TARGET_GCC
#pragma GCC diagnostic push
#pragma GCC diagnostic error "-Wswitch"
#endif
switch(format) {
case PixelFormat::R16F:
case PixelFormat::RG16F:
case PixelFormat::RGB16F:
case PixelFormat::RGBA16F:
case PixelFormat::R32F:
case PixelFormat::RG32F:
case PixelFormat::RGB32F:
case PixelFormat::RGBA32F:
return true;
case PixelFormat::R8Unorm:
case PixelFormat::RG8Unorm:
case PixelFormat::RGB8Unorm:
case PixelFormat::RGBA8Unorm:
case PixelFormat::R8Snorm:
case PixelFormat::RG8Snorm:
case PixelFormat::RGB8Snorm:
case PixelFormat::RGBA8Snorm:
case PixelFormat::R8Srgb:
case PixelFormat::RG8Srgb:
case PixelFormat::RGB8Srgb:
case PixelFormat::RGBA8Srgb:
case PixelFormat::R8UI:
case PixelFormat::RG8UI:
case PixelFormat::RGB8UI:
case PixelFormat::RGBA8UI:
case PixelFormat::R8I:
case PixelFormat::RG8I:
case PixelFormat::RGB8I:
case PixelFormat::RGBA8I:
case PixelFormat::R16Unorm:
case PixelFormat::RG16Unorm:
case PixelFormat::RGB16Unorm:
case PixelFormat::RGBA16Unorm:
case PixelFormat::R16Snorm:
case PixelFormat::RG16Snorm:
case PixelFormat::RGB16Snorm:
case PixelFormat::RGBA16Snorm:
case PixelFormat::R16UI:
case PixelFormat::RG16UI:
case PixelFormat::RGB16UI:
case PixelFormat::RGBA16UI:
case PixelFormat::R16I:
case PixelFormat::RG16I:
case PixelFormat::RGB16I:
case PixelFormat::RGBA16I:
case PixelFormat::R32UI:
case PixelFormat::RG32UI:
case PixelFormat::RGB32UI:
case PixelFormat::RGBA32UI:
case PixelFormat::R32I:
case PixelFormat::RG32I:
case PixelFormat::RGB32I:
case PixelFormat::RGBA32I:
return false;
case PixelFormat::Depth16Unorm:
case PixelFormat::Depth24Unorm:
case PixelFormat::Depth32F:
case PixelFormat::Stencil8UI:
case PixelFormat::Depth16UnormStencil8UI:
case PixelFormat::Depth24UnormStencil8UI:
case PixelFormat::Depth32FStencil8UI:
CORRADE_ASSERT_UNREACHABLE("isPixelFormatFloatingPoint(): can't determine type of" << format, {});
}
#ifdef CORRADE_TARGET_GCC
#pragma GCC diagnostic pop
#endif
CORRADE_ASSERT_UNREACHABLE("isPixelFormatFloatingPoint(): invalid format" << format, {});
}
Add PixelFormat decomposition utilities. Similar to what VertexFormat already has -- getting channel count, channel format, sRGB and color vs depth/stencil properties out of a PixelFormat value. Very useful for various image conversion plugins.
4 years ago
bool isPixelFormatSrgb(const PixelFormat format) {
CORRADE_ASSERT(!isPixelFormatImplementationSpecific(format),
"isPixelFormatSrgb(): can't determine colorspace of an implementation-specific format" << reinterpret_cast<void*>(pixelFormatUnwrap(format)), {});
#ifdef CORRADE_TARGET_GCC
#pragma GCC diagnostic push
#pragma GCC diagnostic error "-Wswitch"
#endif
switch(format) {
case PixelFormat::R8Srgb:
case PixelFormat::RG8Srgb:
case PixelFormat::RGB8Srgb:
case PixelFormat::RGBA8Srgb:
return true;
case PixelFormat::R8Unorm:
case PixelFormat::RG8Unorm:
case PixelFormat::RGB8Unorm:
case PixelFormat::RGBA8Unorm:
case PixelFormat::R8Snorm:
case PixelFormat::RG8Snorm:
case PixelFormat::RGB8Snorm:
case PixelFormat::RGBA8Snorm:
case PixelFormat::R8UI:
case PixelFormat::RG8UI:
case PixelFormat::RGB8UI:
case PixelFormat::RGBA8UI:
case PixelFormat::R8I:
case PixelFormat::RG8I:
case PixelFormat::RGB8I:
case PixelFormat::RGBA8I:
case PixelFormat::R16Unorm:
case PixelFormat::RG16Unorm:
case PixelFormat::RGB16Unorm:
case PixelFormat::RGBA16Unorm:
case PixelFormat::R16Snorm:
case PixelFormat::RG16Snorm:
case PixelFormat::RGB16Snorm:
case PixelFormat::RGBA16Snorm:
case PixelFormat::R16UI:
case PixelFormat::RG16UI:
case PixelFormat::RGB16UI:
case PixelFormat::RGBA16UI:
case PixelFormat::R16I:
case PixelFormat::RG16I:
case PixelFormat::RGB16I:
case PixelFormat::RGBA16I:
case PixelFormat::R32UI:
case PixelFormat::RG32UI:
case PixelFormat::RGB32UI:
case PixelFormat::RGBA32UI:
case PixelFormat::R32I:
case PixelFormat::RG32I:
case PixelFormat::RGB32I:
case PixelFormat::RGBA32I:
case PixelFormat::R16F:
case PixelFormat::RG16F:
case PixelFormat::RGB16F:
case PixelFormat::RGBA16F:
case PixelFormat::R32F:
case PixelFormat::RG32F:
case PixelFormat::RGB32F:
case PixelFormat::RGBA32F:
It makes no sense to call isPixelFormatSrgb() on a depth/stencil format. Doing so usually points to some error in user code, so directly disallow it, instead of returning false. Such behavior is consistent with pixelFormatChannelCount() and other helper APIs.
3 years ago
return false;
Add PixelFormat decomposition utilities. Similar to what VertexFormat already has -- getting channel count, channel format, sRGB and color vs depth/stencil properties out of a PixelFormat value. Very useful for various image conversion plugins.
4 years ago
case PixelFormat::Depth16Unorm:
case PixelFormat::Depth24Unorm:
case PixelFormat::Depth32F:
case PixelFormat::Stencil8UI:
case PixelFormat::Depth16UnormStencil8UI:
case PixelFormat::Depth24UnormStencil8UI:
case PixelFormat::Depth32FStencil8UI:
It makes no sense to call isPixelFormatSrgb() on a depth/stencil format. Doing so usually points to some error in user code, so directly disallow it, instead of returning false. Such behavior is consistent with pixelFormatChannelCount() and other helper APIs.
3 years ago
CORRADE_ASSERT_UNREACHABLE("isPixelFormatSrgb(): can't determine colorspace of" << format, {});
Add PixelFormat decomposition utilities. Similar to what VertexFormat already has -- getting channel count, channel format, sRGB and color vs depth/stencil properties out of a PixelFormat value. Very useful for various image conversion plugins.
4 years ago
}
#ifdef CORRADE_TARGET_GCC
#pragma GCC diagnostic pop
#endif
CORRADE_ASSERT_UNREACHABLE("isPixelFormatSrgb(): invalid format" << format, {});
}
bool isPixelFormatDepthOrStencil(const PixelFormat format) {
CORRADE_ASSERT(!isPixelFormatImplementationSpecific(format),
"isPixelFormatDepthOrStencil(): can't determine type of an implementation-specific format" << reinterpret_cast<void*>(pixelFormatUnwrap(format)), {});
#ifdef CORRADE_TARGET_GCC
#pragma GCC diagnostic push
#pragma GCC diagnostic error "-Wswitch"
#endif
switch(format) {
case PixelFormat::R8Unorm:
case PixelFormat::RG8Unorm:
case PixelFormat::RGB8Unorm:
case PixelFormat::RGBA8Unorm:
case PixelFormat::R8Snorm:
case PixelFormat::RG8Snorm:
case PixelFormat::RGB8Snorm:
case PixelFormat::RGBA8Snorm:
case PixelFormat::R8Srgb:
case PixelFormat::RG8Srgb:
case PixelFormat::RGB8Srgb:
case PixelFormat::RGBA8Srgb:
case PixelFormat::R8UI:
case PixelFormat::RG8UI:
case PixelFormat::RGB8UI:
case PixelFormat::RGBA8UI:
case PixelFormat::R8I:
case PixelFormat::RG8I:
case PixelFormat::RGB8I:
case PixelFormat::RGBA8I:
case PixelFormat::R16Unorm:
case PixelFormat::RG16Unorm:
case PixelFormat::RGB16Unorm:
case PixelFormat::RGBA16Unorm:
case PixelFormat::R16Snorm:
case PixelFormat::RG16Snorm:
case PixelFormat::RGB16Snorm:
case PixelFormat::RGBA16Snorm:
case PixelFormat::R16UI:
case PixelFormat::RG16UI:
case PixelFormat::RGB16UI:
case PixelFormat::RGBA16UI:
case PixelFormat::R16I:
case PixelFormat::RG16I:
case PixelFormat::RGB16I:
case PixelFormat::RGBA16I:
case PixelFormat::R32UI:
case PixelFormat::RG32UI:
case PixelFormat::RGB32UI:
case PixelFormat::RGBA32UI:
case PixelFormat::R32I:
case PixelFormat::RG32I:
case PixelFormat::RGB32I:
case PixelFormat::RGBA32I:
case PixelFormat::R16F:
case PixelFormat::RG16F:
case PixelFormat::RGB16F:
case PixelFormat::RGBA16F:
case PixelFormat::R32F:
case PixelFormat::RG32F:
case PixelFormat::RGB32F:
case PixelFormat::RGBA32F:
return false;
case PixelFormat::Depth16Unorm:
case PixelFormat::Depth24Unorm:
case PixelFormat::Depth32F:
case PixelFormat::Stencil8UI:
case PixelFormat::Depth16UnormStencil8UI:
case PixelFormat::Depth24UnormStencil8UI:
case PixelFormat::Depth32FStencil8UI:
return true;
}
#ifdef CORRADE_TARGET_GCC
#pragma GCC diagnostic pop
#endif
CORRADE_ASSERT_UNREACHABLE("isPixelFormatDepthOrStencil(): invalid format" << format, {});
}
Add a pixelFormat(PixelFormat, UnsignedInt, bool) helper. Useful for creating pixel formats with different channel count, adding/removing the sRGB bit and such. Counterpart to vertexFormat(VertexFormat, UnsignedInt, bool) that got added back in 2020.06 already.
3 years ago
PixelFormat pixelFormat(const PixelFormat format, const UnsignedInt channelCount, const bool srgb) {
CORRADE_ASSERT(!isPixelFormatImplementationSpecific(format),
"pixelFormat(): can't assemble a format out of an implementation-specific format" << reinterpret_cast<void*>(pixelFormatUnwrap(format)), {});
CORRADE_ASSERT(!isPixelFormatDepthOrStencil(format),
"pixelFormat(): can't assemble a format out of" << format, {});
PixelFormat channelFormat = pixelFormatChannelFormat(format);
/* First turn the format into a sRGB one or remove the sRGB property, if
requested. The [RGBA]8Srgb formats follow [RGBA]8Unorm in the same order
so it's just constant addition / subtraction for all four variants. */
if(srgb && channelFormat != PixelFormat::R8Srgb) {
CORRADE_ASSERT(channelFormat == PixelFormat::R8Unorm,
"pixelFormat():" << format << "can't be made sRGB", {});
channelFormat = PixelFormat(UnsignedInt(channelFormat) - UnsignedInt(PixelFormat::R8Unorm) + UnsignedInt(PixelFormat::R8Srgb));
} else if(!srgb && channelFormat == PixelFormat::R8Srgb) {
channelFormat = PixelFormat(UnsignedInt(channelFormat) - UnsignedInt(PixelFormat::R8Srgb) + UnsignedInt(PixelFormat::R8Unorm));
}
CORRADE_ASSERT(channelCount >= 1 && channelCount <= 4,
"pixelFormat(): invalid component count" << channelCount, {});
/* The two-, three- and four-channel variants follow each other, so it's
just addition again. There may be packed formats in the future, so
whitelist for the known set of single-channel formats. */
if(channelFormat == PixelFormat::R8Unorm ||
channelFormat == PixelFormat::R8Snorm ||
channelFormat == PixelFormat::R8Srgb ||
channelFormat == PixelFormat::R8UI ||
channelFormat == PixelFormat::R8I ||
channelFormat == PixelFormat::R16Unorm ||
channelFormat == PixelFormat::R16Snorm ||
channelFormat == PixelFormat::R16UI ||
channelFormat == PixelFormat::R16I ||
channelFormat == PixelFormat::R32UI ||
channelFormat == PixelFormat::R32I ||
channelFormat == PixelFormat::R16F ||
channelFormat == PixelFormat::R32F)
return PixelFormat(UnsignedInt(channelFormat) + channelCount - 1);
CORRADE_INTERNAL_ASSERT_UNREACHABLE(); /* LCOV_EXCL_LINE */
}
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
namespace {
#ifndef DOXYGEN_GENERATING_OUTPUT /* It gets *really* confused */
constexpr const char* PixelFormatNames[] {
#define _c(format) #format,
#include "Magnum/Implementation/pixelFormatMapping.hpp"
#undef _c
};
#endif
}
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
#ifndef DOXYGEN_GENERATING_OUTPUT
Debug& operator<<(Debug& debug, const PixelFormat value) {
Packed debug output for enums used by Trade. Again not publicly documented because I don't like the naming and I don't have the full behavior and interactions figured out yet -- i.e., an array of VertexFormats would be printed with Debug::packed as a long string of characters without any whitespace. Not good, thus this feature probably needs to be split in two, with this being named "compact" or something else.
4 years ago
const bool packed = debug.immediateFlags() >= Debug::Flag::Packed;
if(!packed)
debug << "PixelFormat" << Debug::nospace;
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
if(isPixelFormatImplementationSpecific(value)) {
Packed debug output for enums used by Trade. Again not publicly documented because I don't like the naming and I don't have the full behavior and interactions figured out yet -- i.e., an array of VertexFormats would be printed with Debug::packed as a long string of characters without any whitespace. Not good, thus this feature probably needs to be split in two, with this being named "compact" or something else.
4 years ago
return debug << (packed ? "ImplementationSpecific(" : "::ImplementationSpecific(") << Debug::nospace << reinterpret_cast<void*>(pixelFormatUnwrap(value)) << Debug::nospace << ")";
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
}
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
if(UnsignedInt(value) - 1 < Containers::arraySize(PixelFormatNames)) {
Packed debug output for enums used by Trade. Again not publicly documented because I don't like the naming and I don't have the full behavior and interactions figured out yet -- i.e., an array of VertexFormats would be printed with Debug::packed as a long string of characters without any whitespace. Not good, thus this feature probably needs to be split in two, with this being named "compact" or something else.
4 years ago
return debug << (packed ? "" : "::") << Debug::nospace << PixelFormatNames[UnsignedInt(value) - 1];
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
}
Packed debug output for enums used by Trade. Again not publicly documented because I don't like the naming and I don't have the full behavior and interactions figured out yet -- i.e., an array of VertexFormats would be printed with Debug::packed as a long string of characters without any whitespace. Not good, thus this feature probably needs to be split in two, with this being named "compact" or something else.
4 years ago
return debug << (packed ? "" : "(") << Debug::nospace << reinterpret_cast<void*>(UnsignedInt(value)) << Debug::nospace << (packed ? "" : ")");
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
}
#endif
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
namespace {
#ifndef DOXYGEN_GENERATING_OUTPUT /* It gets *really* confused */
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
constexpr const char* CompressedPixelFormatNames[] {
#define _c(format, width, height, depth, size) #format,
#include "Magnum/Implementation/compressedPixelFormatMapping.hpp"
#undef _c
};
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
constexpr UnsignedShort CompressedBlockData[] {
/* Assuming w/h/d/s is never larger than 16 (and never zero), each number
has 1 subtracted and packed into four bits, 16 bits in total. The size
is supplied in bits, so first divide by eight and then subtract 1. For
Add HDR ASTC pixel formats. Both Vulkan and Metal have separate enums for these, GL is the only odd one out.
7 years ago
the currently ~100 supported formats that makes this table to be about
256 bytes.*/
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
#define _c(format, width, height, depth, size) \
((width - 1) << 12) | \
((height - 1) << 8) | \
((depth - 1) << 4) | \
((size >> 3) - 1),
#include "Magnum/Implementation/compressedPixelFormatMapping.hpp"
#undef _c
};
#endif
}
Rename pixelSize() and other helpers to pixelFormatSize() etc. For consistency with how VertexFormat and other enum helpers are named. The compressedBlockSize() and compressedBlockDataSize() is also renamed to compressedPixelFormatBlockSize() and compressedPixelFormatBlockDataSize(). While backwards compatibility aliases are in place, a breaking change is that Image classes now look for pixelFormatSize() instead of pixelSize(). This is used e.g. when passing GL::PixelFormat / GL::PixelType to the image classes, instead of the generic PixelFormat. While useful, it's unlikely that any project was defining their own pixel format enum and pixelSize() for a D3D or Metal renderer or whatnot, so the breakage should have no practical impact.
4 years ago
Vector3i compressedPixelFormatBlockSize(const CompressedPixelFormat format) {
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
CORRADE_ASSERT(!(UnsignedInt(format) & (1 << 31)),
Rename pixelSize() and other helpers to pixelFormatSize() etc. For consistency with how VertexFormat and other enum helpers are named. The compressedBlockSize() and compressedBlockDataSize() is also renamed to compressedPixelFormatBlockSize() and compressedPixelFormatBlockDataSize(). While backwards compatibility aliases are in place, a breaking change is that Image classes now look for pixelFormatSize() instead of pixelSize(). This is used e.g. when passing GL::PixelFormat / GL::PixelType to the image classes, instead of the generic PixelFormat. While useful, it's unlikely that any project was defining their own pixel format enum and pixelSize() for a D3D or Metal renderer or whatnot, so the breakage should have no practical impact.
4 years ago
"compressedPixelFormatBlockSize(): can't determine size of an implementation-specific format" << reinterpret_cast<void*>(compressedPixelFormatUnwrap(format)), {});
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
CORRADE_ASSERT(UnsignedInt(format) - 1 < Containers::arraySize(CompressedBlockData),
Rename pixelSize() and other helpers to pixelFormatSize() etc. For consistency with how VertexFormat and other enum helpers are named. The compressedBlockSize() and compressedBlockDataSize() is also renamed to compressedPixelFormatBlockSize() and compressedPixelFormatBlockDataSize(). While backwards compatibility aliases are in place, a breaking change is that Image classes now look for pixelFormatSize() instead of pixelSize(). This is used e.g. when passing GL::PixelFormat / GL::PixelType to the image classes, instead of the generic PixelFormat. While useful, it's unlikely that any project was defining their own pixel format enum and pixelSize() for a D3D or Metal renderer or whatnot, so the breakage should have no practical impact.
4 years ago
"compressedPixelFormatBlockSize(): invalid format" << format, {});
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
const UnsignedInt data = CompressedBlockData[UnsignedInt(format) - 1];
return {
(Int(data >> 12) & 0xf) + 1,
(Int(data >> 8) & 0xf) + 1,
(Int(data >> 4) & 0xf) + 1,
};
}
Rename pixelSize() and other helpers to pixelFormatSize() etc. For consistency with how VertexFormat and other enum helpers are named. The compressedBlockSize() and compressedBlockDataSize() is also renamed to compressedPixelFormatBlockSize() and compressedPixelFormatBlockDataSize(). While backwards compatibility aliases are in place, a breaking change is that Image classes now look for pixelFormatSize() instead of pixelSize(). This is used e.g. when passing GL::PixelFormat / GL::PixelType to the image classes, instead of the generic PixelFormat. While useful, it's unlikely that any project was defining their own pixel format enum and pixelSize() for a D3D or Metal renderer or whatnot, so the breakage should have no practical impact.
4 years ago
#ifdef MAGNUM_BUILD_DEPRECATED
Vector3i compressedBlockSize(const CompressedPixelFormat format) {
return compressedPixelFormatBlockSize(format);
}
#endif
UnsignedInt compressedPixelFormatBlockDataSize(const CompressedPixelFormat format) {
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
CORRADE_ASSERT(!(UnsignedInt(format) & (1 << 31)),
Rename pixelSize() and other helpers to pixelFormatSize() etc. For consistency with how VertexFormat and other enum helpers are named. The compressedBlockSize() and compressedBlockDataSize() is also renamed to compressedPixelFormatBlockSize() and compressedPixelFormatBlockDataSize(). While backwards compatibility aliases are in place, a breaking change is that Image classes now look for pixelFormatSize() instead of pixelSize(). This is used e.g. when passing GL::PixelFormat / GL::PixelType to the image classes, instead of the generic PixelFormat. While useful, it's unlikely that any project was defining their own pixel format enum and pixelSize() for a D3D or Metal renderer or whatnot, so the breakage should have no practical impact.
4 years ago
"compressedPixelFormatBlockDataSize(): can't determine size of an implementation-specific format" << reinterpret_cast<void*>(compressedPixelFormatUnwrap(format)), {});
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
CORRADE_ASSERT(UnsignedInt(format) - 1 < Containers::arraySize(CompressedBlockData),
Rename pixelSize() and other helpers to pixelFormatSize() etc. For consistency with how VertexFormat and other enum helpers are named. The compressedBlockSize() and compressedBlockDataSize() is also renamed to compressedPixelFormatBlockSize() and compressedPixelFormatBlockDataSize(). While backwards compatibility aliases are in place, a breaking change is that Image classes now look for pixelFormatSize() instead of pixelSize(). This is used e.g. when passing GL::PixelFormat / GL::PixelType to the image classes, instead of the generic PixelFormat. While useful, it's unlikely that any project was defining their own pixel format enum and pixelSize() for a D3D or Metal renderer or whatnot, so the breakage should have no practical impact.
4 years ago
"compressedPixelFormatBlockDataSize(): invalid format" << format, {});
Add a table describing block size of all known compressed pixel formats. Needed for a bunch of features (and feature parity in CompressedImageView).
7 years ago
return (CompressedBlockData[UnsignedInt(format) - 1] & 0xf) + 1;
}
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
#ifndef DOXYGEN_GENERATING_OUTPUT
Debug& operator<<(Debug& debug, const CompressedPixelFormat value) {
Packed debug output for enums used by Trade. Again not publicly documented because I don't like the naming and I don't have the full behavior and interactions figured out yet -- i.e., an array of VertexFormats would be printed with Debug::packed as a long string of characters without any whitespace. Not good, thus this feature probably needs to be split in two, with this being named "compact" or something else.
4 years ago
const bool packed = debug.immediateFlags() >= Debug::Flag::Packed;
if(!packed)
debug << "CompressedPixelFormat" << Debug::nospace;
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
if(isCompressedPixelFormatImplementationSpecific(value)) {
Packed debug output for enums used by Trade. Again not publicly documented because I don't like the naming and I don't have the full behavior and interactions figured out yet -- i.e., an array of VertexFormats would be printed with Debug::packed as a long string of characters without any whitespace. Not good, thus this feature probably needs to be split in two, with this being named "compact" or something else.
4 years ago
return debug << (packed ? "ImplementationSpecific(" : "::ImplementationSpecific(") << Debug::nospace << reinterpret_cast<void*>(compressedPixelFormatUnwrap(value)) << Debug::nospace << ")";
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
}
Optimize {Compressed,}PixelFormat debug output. Instead of prefixing every value with (Compressed)PixelFormat::, printing that just once. Also moved the name table to an external header so it can be used later for configuration value parsing. Compared to the previous commit, Bloaty reports a save, meaning these two commits in total save about 4 kB: VM SIZE FILE SIZE -------------- -------------- +687% +3.70Ki .rela.dyn +3.70Ki +687% +15e2% +1.26Ki .data.rel.ro +1.26Ki +15e2% [ = ] 0 .strtab +100 +0.1% [ = ] 0 .symtab +48 +0.2% -50.0% -8 [LOAD [RW]] -8 -50.0% -3.1% -888 .eh_frame -888 -3.1% -1.5% -1.92Ki .text -1.92Ki -1.5% [ = ] 0 [Unmapped] -1.97Ki -32.2% -38.5% -4.19Ki .rodata -4.19Ki -38.5% -0.7% -2.02Ki TOTAL -3.85Ki -1.0%
7 years ago
if(UnsignedInt(value) - 1 < Containers::arraySize(CompressedPixelFormatNames)) {
Packed debug output for enums used by Trade. Again not publicly documented because I don't like the naming and I don't have the full behavior and interactions figured out yet -- i.e., an array of VertexFormats would be printed with Debug::packed as a long string of characters without any whitespace. Not good, thus this feature probably needs to be split in two, with this being named "compact" or something else.
4 years ago
return debug << (packed ? "" : "::") << Debug::nospace << CompressedPixelFormatNames[UnsignedInt(value) - 1];
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
}
Packed debug output for enums used by Trade. Again not publicly documented because I don't like the naming and I don't have the full behavior and interactions figured out yet -- i.e., an array of VertexFormats would be printed with Debug::packed as a long string of characters without any whitespace. Not good, thus this feature probably needs to be split in two, with this being named "compact" or something else.
4 years ago
return debug << (packed ? "" : "(") << Debug::nospace << reinterpret_cast<void*>(UnsignedInt(value)) << Debug::nospace << (packed ? "" : ")");
Split the OpenGL layer out, pt 9: generic pixel formats. This is quite big, so: * There are new Magnum::PixelFormat and Magnum::CompressedPixelFormat enums, which contain generic API-independent formats. In particular, PixelFormat replaces GL::PixelFormat and GL::PixelType with a single value. * There's GL::pixelFormat(), GL::pixelType(), GL::compressedPixelFormat() to convert the generic enums to GL-specific. The mapping is only in one direction, done with a lookup table (generic enums are indices to that table). * GL classes taking the formats directly (such as GL::BufferImage) have overloads that take both the GL-specific and generic format. * The generic Image, CompressedImage, ImageView, CompressedImageView, and Trade::ImageData classes now accept the generic formats first-class. However, it's also possible to store an implementation-specific value to cover cases where a generic format enum doesn't have support for a particular format. This is done by wrapping the value using pixelFormatWrap() or compressedPixelFormatWrap(). Particular GPU APIs then assume it's their implementation-specific value and extract the value back using pixelFormatUnwrap() or compressedPixelFormatUnwrap(). There's also an isPixelFormatImplementationSpecific() and isCompressedPixelFormatImplementationSpecific() that distinguishes these values. * Many operations need pixel size and in order to have it even for implementation-specific formats, a corresponding pixelSize() overload is found via ADL on construction and the calculated size stored along the format. Previously the pixel size was only calculated on demand, but that's not possible now. In case such overload is not available, it's possible to pass pixel size manually as well. * In order to support the GL format+type pair, Image, ImageView and Trade::ImageData, there's now an additional untyped formatExtra() field that holds the second value. * The CompressedPixelStorage class is now unconditionally available on all targets, including OpenGL ES and WebGL. However, on OpenGL ES the GL APIs expect that it's all at default values. I attempted to preserve backwards compatibility as much as possible: * The PixelFormat and CompressedPixelFormat enum now contains generic API-independent values. The GL-specific formats are present there, but marked as deprecated. Use either the generic values or GL::PixelFormat (togehter with GL::PixelType) and GL::CompressedPixelFormat instead. There's a lot of ugliness caused by this, but seems to work well. * *Image::type() functions are deprecated as they were too GL-specific. Use formatExtra() and cast it to GL::PixelType instead. * Image constructors take templated format or format+extra arguments, so passing GL-specific values to them should still work.
8 years ago
}
#endif
}
Make (Compressed)PixelFormat work with Configuration. And thus also Arguments.
7 years ago
namespace Corrade { namespace Utility {
Use Corrade String[View] in some ConfigurationValue specializations. These headers are included quite often and because <Corrade/Utility/StlForwardString.h> expands to the whole <string>, it has quite a significant impact. Other cases like Math/ConfigurationValue.h are not changed at the moment, as that header is only included together with Utility::Arguments or Utility::Configuration -- and these classes are still STL-heavy, so avoiding std::string usage in just the ConfigurationValue specializations doesn't help anything at the moment.
4 years ago
Containers::String ConfigurationValue<Magnum::PixelFormat>::toString(Magnum::PixelFormat value, ConfigurationValueFlags) {
Make (Compressed)PixelFormat work with Configuration. And thus also Arguments.
7 years ago
if(Magnum::UnsignedInt(value) - 1 < Containers::arraySize(Magnum::PixelFormatNames))
return Magnum::PixelFormatNames[Magnum::UnsignedInt(value) - 1];
return {};
}
Use Corrade String[View] in some ConfigurationValue specializations. These headers are included quite often and because <Corrade/Utility/StlForwardString.h> expands to the whole <string>, it has quite a significant impact. Other cases like Math/ConfigurationValue.h are not changed at the moment, as that header is only included together with Utility::Arguments or Utility::Configuration -- and these classes are still STL-heavy, so avoiding std::string usage in just the ConfigurationValue specializations doesn't help anything at the moment.
4 years ago
Magnum::PixelFormat ConfigurationValue<Magnum::PixelFormat>::fromString(Containers::StringView stringValue, ConfigurationValueFlags) {
Make (Compressed)PixelFormat work with Configuration. And thus also Arguments.
7 years ago
/** @todo This is extremely slow with >100 values. Do a binary search on a
sorted index list instead (extracted into a common utility) */
for(std::size_t i = 0; i != Containers::arraySize(Magnum::PixelFormatNames); ++i)
if(stringValue == Magnum::PixelFormatNames[i]) return Magnum::PixelFormat(i + 1);
return {};
}
Use Corrade String[View] in some ConfigurationValue specializations. These headers are included quite often and because <Corrade/Utility/StlForwardString.h> expands to the whole <string>, it has quite a significant impact. Other cases like Math/ConfigurationValue.h are not changed at the moment, as that header is only included together with Utility::Arguments or Utility::Configuration -- and these classes are still STL-heavy, so avoiding std::string usage in just the ConfigurationValue specializations doesn't help anything at the moment.
4 years ago
Containers::String ConfigurationValue<Magnum::CompressedPixelFormat>::toString(Magnum::CompressedPixelFormat value, ConfigurationValueFlags) {
Make (Compressed)PixelFormat work with Configuration. And thus also Arguments.
7 years ago
if(Magnum::UnsignedInt(value) - 1 < Containers::arraySize(Magnum::CompressedPixelFormatNames))
return Magnum::CompressedPixelFormatNames[Magnum::UnsignedInt(value) - 1];
return {};
}
Use Corrade String[View] in some ConfigurationValue specializations. These headers are included quite often and because <Corrade/Utility/StlForwardString.h> expands to the whole <string>, it has quite a significant impact. Other cases like Math/ConfigurationValue.h are not changed at the moment, as that header is only included together with Utility::Arguments or Utility::Configuration -- and these classes are still STL-heavy, so avoiding std::string usage in just the ConfigurationValue specializations doesn't help anything at the moment.
4 years ago
Magnum::CompressedPixelFormat ConfigurationValue<Magnum::CompressedPixelFormat>::fromString(Containers::StringView stringValue, ConfigurationValueFlags) {
Make (Compressed)PixelFormat work with Configuration. And thus also Arguments.
7 years ago
/** @todo This is extremely slow with >100 values. Do a binary search on a
sorted index list instead (extracted into a common utility) */
for(std::size_t i = 0; i != Containers::arraySize(Magnum::CompressedPixelFormatNames); ++i)
if(stringValue == Magnum::CompressedPixelFormatNames[i]) return Magnum::CompressedPixelFormat(i + 1);
return {};
}
}}