qianqians / gli Goto Github PK

OpenGL Image (GLI) is a header only C++ image library for graphics software.

GLI provides classes and functions to load image files (KTX and DDS), facilitate graphics APIs texture creation, compare textures, access texture texels, sample textures, convert textures, generate mipmaps, etc.

This library works perfectly with OpenGL or Vulkan but it also ensures interoperability with other third party libraries and SDK. It is a good candidate for software rendering (raytracing / rasterisation), image processing, image based software testing or any development context that requires a simple and convenient image library.

GLI is written in C++11. It is a platform independent library with no dependence and it supports the following compilers:

• Apple Clang 6.0 and higher
• GCC 4.7 and higher
• Intel C++ Composer XE 2013 and higher
• LLVM 3.4 and higher
• Visual C++ 2013 and higher
• Any C++11 compiler

For more information about GLI, please have a look at the manual and the API reference documentation. The source code and the documentation are licensed under the Happy Bunny License (Modified MIT) or the MIT License.

Thanks for contributing to the project by submitting pull requests.

#include <gli/gli.hpp>

GLuint CreateTexture(char const* Filename)
{
	gli::texture Texture = gli::load(Filename);
	if(Texture.empty())
		return 0;

	gli::gl GL(gli::gl::PROFILE_GL33);
	gli::gl::format const Format = GL.translate(Texture.format(), Texture.swizzles());
	GLenum Target = GL.translate(Texture.target());
	assert(gli::is_compressed(Texture.format()) && Target == gli::TARGET_2D);

	GLuint TextureName = 0;
	glGenTextures(1, &TextureName);
	glBindTexture(Target, TextureName);
	glTexParameteri(Target, GL_TEXTURE_BASE_LEVEL, 0);
	glTexParameteri(Target, GL_TEXTURE_MAX_LEVEL, static_cast<GLint>(Texture.levels() - 1));
	glTexParameteriv(Target, GL_TEXTURE_SWIZZLE_RGBA, &Format.Swizzles[0]);
	glTexStorage2D(Target, static_cast<GLint>(Texture.levels()), Format.Internal, Extent.x, Extent.y);
	for(std::size_t Level = 0; Level < Texture.levels(); ++Level)
	{
		glm::tvec3<GLsizei> Extent(Texture.extent(Level));
		glCompressedTexSubImage2D(
			Target, static_cast<GLint>(Level), 0, 0, Extent.x, Extent.y,
			Format.Internal, static_cast<GLsizei>(Texture.size(Level)), Texture.data(0, 0, Level));
	}

	return TextureName;
}

Service	System	Compiler	Status
Travis CI	Linux 64 bits	Clang 3.9, Clang 7, Clang 9, Clang 10, GCC 4.8, GCC 7.4, GCC 9, GCC 10
AppVeyor	Windows 32 and 64	Visual Studio 2013

• Added decompression and sampling of DXT1, DXT3, DXT5, ATI1N and ATI2N #110 #119
• Added depth and stencil format queries #119
• Added texture_grad to samplers

• Fixed R8 SRGB #120

• Extend flip() for S3TC compressed textures #94
• Added format property queries #102

• Fixed texture operator=
• Added initial manual

• Fixed ATI2N swizzle parameters #121

• Updated GLM to 0.9.8.0 release

• Fixed KTX cube maps saving
• Fixed texture::clear build

• Added texture copy, no allocation involved, only transfer
• Added sub-image copy
• Added non-member clear
• Added make_texture* helper functions

• Added compressed npot textures support #73
• Added image access cache to generic textures
• Added luminance alpha format translation to OpenGL 3.3+ through RG swizzling. #93
• Improved performance when sampling using mipmaps
• Improved nearest filter without border using texture_lod performance (~2.5x faster with texture2d)
• Improved texture::data() (~18x with cube array, ~68x with 2D)
• Improved texture::size() (~3.1x with cube array, ~3.9x with 2D)
• Improved simultanous texture::size() and texture::extent() calls (~2.1x with cube array, ~2.3x with 2D)

• Fixed DX10 DDS saving of 3D textures #91
• Fixed BGRX translation with PROFILE_GL33 profile #92
• Fixed DDS9 loading of 3D texture detected as 2D texture #93

• Added texture swizzle support #79
• Added texture memory swizzle support
• Added texture conversion from any uncompressed format to any uncompressed format
• Added texture lod
• Added texture mipmaps generation for uncompressed formats
• Added support for load and store
• Added support for many new formats
• Added sampler 1D, 2D, 3D, array and cube map
• Added sampler texel fetch and texel write
• Added sampler clear
• Added transform algorithm to compute arithmetic between texels
• Added reduce algorithm to compare all texels within an image

• Reordered formats to match Vulkan formats
• Improved OpenGL translation with for multiple profiles: KTX, ES2.0, ES3.0, GL3.2 and GL3.3
• Improved Doxygen documentation

• Fixed PVRTC2 support
• Fixed luminance and alpha translation to OpenGL #56
• Fixed DXGI_FORMAT_B8G8R8X8_UNORM_SRGB support #59
• Fixed FORMAT_RGBA8_UNORM DDS loading using DDPF_RGBA mode #60
• Fixed handling of DDS DDPF_ALPHAPIXELS #68
• Fixed images, better matching names and formats #78 #81 #80
• Fixed BC4U and BC5U files generated from MS DDS loader #82

• Added KMG container support and spec proposal

• Added KTX loading and saving
• Added gli::load for generic file loading, either DDS or KTX files depending on filename extensions
• Added gli::save for generic file saving, either DDS or KTX files depending on filename extensions
• Added texture views using different texture format, including compressed texture formats
• Added fine granularity includes
• Improved API documentation
• Much faster texture comparisons is non optimal cases. (Measured ~21x faster on Intel IVB)
• Explicitly handling of texture targets: fixed various cases of cubemap and texture arrays failing to load with DDS
• Fixed GCC build
• Fixed warnings
• Fixed saved DDS header size on #52

• Updated API documentation
• Fixed link error

• Fixed interface inconsistencies
• Improved clear(), data() and size() performance using caching
• Removed internal dependence to std::fstream
• Added FORMAT_BGRX8_UNORM and FORMAT_BGRX8_SRGB support #48, #43
• Improved FORMAT_RGB8_UNORM loading

• Large refactoring
• Added loading DDS from memory
• Added saving DDS to memory
• Improved DDS coverage for R, RG, RGB and RGBA formats
• Added DDS ASTC, PVRTC, ATC and ETC support
• Added DDS alpha, luminance and alpha luminance support
• Added PVRTC2, ETC2 and EAC formats

• Fixed swizzled RGB channel when reading back a DDS
• Fixed getMask* link error

• Added flip function
• Added level_count function
• Fixed interaction with std::map (#33)
• Added texelFetch and texelWrite functions

• Essencially a rewrite of the library
• Added explicit copies
• Added single memory allocation per texture storage
• Added texture views
• Added texture copies
• Added comparison operators
• Added clear

• Added DDS saving
• Fixed GCC build
• Fixed XCode build

• Large API refactoring
• Performance improvements at loading: 50x in debug and 50% in release build
• Added texture2DArray
• Added textureCube and textureCubeArray
• Added texture3D
• Added texture1D and texture1DArray
• Improved DDS loading support

• Fixed bugs

• Fixed bugs

• Fixed bugs

• Added DDS10 load and store (BC1 - BC7)
• Added extension system
• Added automatic OpenGL texture object creation from file

• Added DDS exporter

• Fixed GCC build

• Added texelWrite function
• Fixed Visual Studio 2010 warnings
• Added readme.txt and copying.txt

• Updated API
• Removed Boost dependency

• Fixed DDS loader
• Added RGB8 to DDS loader
• Added component swizzle operation
• Added 32 bits integer components support
• Fixed texel fetch

• Added TGA loader
• Added DDS loader
• Added duplicate, crop, partial copy
• Added mipmaps generation.