From 8d7e5e422474b4496adbfbdbbe59cf870bec93c9 Mon Sep 17 00:00:00 2001 From: Noel Berry Date: Wed, 26 Aug 2020 00:38:01 -0700 Subject: [PATCH] first true commit --- .gitignore | 6 + CMakeLists.txt | 128 + README.md | 0 private/blah/internal/graphics.h | 67 + private/blah/internal/graphics_opengl.cpp | 1630 +++++ private/blah/internal/input.h | 57 + private/blah/internal/platform.h | 120 + private/blah/internal/platform_sdl2.cpp | 652 ++ private/blah/third_party/stb_image.h | 7656 ++++++++++++++++++++ private/blah/third_party/stb_image_write.h | 1666 +++++ private/blah/third_party/stb_truetype.h | 5011 +++++++++++++ public/blah.h | 52 + public/blah/app.cpp | 238 + public/blah/app.h | 73 + public/blah/containers/list.h | 314 + public/blah/containers/stacklist.h | 212 + public/blah/containers/str.cpp | 385 + public/blah/containers/str.h | 185 + public/blah/drawing/batch.cpp | 832 +++ public/blah/drawing/batch.h | 169 + public/blah/drawing/spritefont.cpp | 250 + public/blah/drawing/spritefont.h | 93 + public/blah/drawing/subtexture.cpp | 41 + public/blah/drawing/subtexture.h | 24 + public/blah/filesystem.cpp | 136 + public/blah/filesystem.h | 41 + public/blah/graphics/framebuffer.h | 36 + public/blah/graphics/graphics.cpp | 340 + public/blah/graphics/graphics.h | 349 + public/blah/graphics/material.cpp | 252 + public/blah/graphics/material.h | 66 + public/blah/graphics/mesh.h | 47 + public/blah/graphics/shader.h | 36 + public/blah/graphics/texture.h | 57 + public/blah/images/aseprite.cpp | 466 ++ public/blah/images/aseprite.h | 152 + public/blah/images/font.cpp | 244 + public/blah/images/font.h | 62 + public/blah/images/image.cpp | 284 + public/blah/images/image.h | 38 + public/blah/images/packer.cpp | 336 + public/blah/images/packer.h | 71 + public/blah/input/input.cpp | 378 + public/blah/input/input.h | 431 ++ public/blah/input/virtual_axis.cpp | 4 + public/blah/input/virtual_axis.h | 84 + public/blah/input/virtual_button.cpp | 4 + public/blah/input/virtual_button.h | 83 + public/blah/input/virtual_stick.cpp | 4 + public/blah/input/virtual_stick.h | 95 + public/blah/log.cpp | 60 + public/blah/log.h | 34 + public/blah/math/calc.cpp | 135 + public/blah/math/calc.h | 71 + public/blah/math/circle.cpp | 11 + public/blah/math/circle.h | 19 + public/blah/math/color.cpp | 156 + public/blah/math/color.h | 64 + public/blah/math/ease.h | 336 + public/blah/math/line.cpp | 157 + public/blah/math/line.h | 30 + public/blah/math/mat3x2.cpp | 187 + public/blah/math/mat3x2.h | 47 + public/blah/math/mat4x4.cpp | 109 + public/blah/math/mat4x4.h | 44 + public/blah/math/point.cpp | 51 + public/blah/math/point.h | 41 + public/blah/math/quad.cpp | 21 + public/blah/math/quad.h | 19 + public/blah/math/rect.cpp | 111 + public/blah/math/rect.h | 101 + public/blah/math/rectI.cpp | 90 + public/blah/math/rectI.h | 64 + public/blah/math/stopwatch.cpp | 26 + public/blah/math/stopwatch.h | 16 + public/blah/math/vec2.cpp | 122 + public/blah/math/vec2.h | 70 + public/blah/math/vec4.h | 15 + public/blah/streams/bufferstream.cpp | 107 + public/blah/streams/bufferstream.h | 37 + public/blah/streams/filestream.cpp | 96 + public/blah/streams/filestream.h | 32 + public/blah/streams/memorystream.cpp | 55 + public/blah/streams/memorystream.h | 35 + public/blah/streams/stream.cpp | 122 + public/blah/streams/stream.h | 108 + public/blah/time.cpp | 54 + public/blah/time.h | 34 + 88 files changed, 26844 insertions(+) create mode 100644 .gitignore create mode 100644 CMakeLists.txt create mode 100644 README.md create mode 100644 private/blah/internal/graphics.h create mode 100644 private/blah/internal/graphics_opengl.cpp create mode 100644 private/blah/internal/input.h create mode 100644 private/blah/internal/platform.h create mode 100644 private/blah/internal/platform_sdl2.cpp create mode 100644 private/blah/third_party/stb_image.h create mode 100644 private/blah/third_party/stb_image_write.h create mode 100644 private/blah/third_party/stb_truetype.h create mode 100644 public/blah.h create mode 100644 public/blah/app.cpp create mode 100644 public/blah/app.h create mode 100644 public/blah/containers/list.h create mode 100644 public/blah/containers/stacklist.h create mode 100644 public/blah/containers/str.cpp create mode 100644 public/blah/containers/str.h create mode 100644 public/blah/drawing/batch.cpp create mode 100644 public/blah/drawing/batch.h create mode 100644 public/blah/drawing/spritefont.cpp create mode 100644 public/blah/drawing/spritefont.h create mode 100644 public/blah/drawing/subtexture.cpp create mode 100644 public/blah/drawing/subtexture.h create mode 100644 public/blah/filesystem.cpp create mode 100644 public/blah/filesystem.h create mode 100644 public/blah/graphics/framebuffer.h create mode 100644 public/blah/graphics/graphics.cpp create mode 100644 public/blah/graphics/graphics.h create mode 100644 public/blah/graphics/material.cpp create mode 100644 public/blah/graphics/material.h create mode 100644 public/blah/graphics/mesh.h create mode 100644 public/blah/graphics/shader.h create mode 100644 public/blah/graphics/texture.h create mode 100644 public/blah/images/aseprite.cpp create mode 100644 public/blah/images/aseprite.h create mode 100644 public/blah/images/font.cpp create mode 100644 public/blah/images/font.h create mode 100644 public/blah/images/image.cpp create mode 100644 public/blah/images/image.h create mode 100644 public/blah/images/packer.cpp create mode 100644 public/blah/images/packer.h create mode 100644 public/blah/input/input.cpp create mode 100644 public/blah/input/input.h create mode 100644 public/blah/input/virtual_axis.cpp create mode 100644 public/blah/input/virtual_axis.h create mode 100644 public/blah/input/virtual_button.cpp create mode 100644 public/blah/input/virtual_button.h create mode 100644 public/blah/input/virtual_stick.cpp create mode 100644 public/blah/input/virtual_stick.h create mode 100644 public/blah/log.cpp create mode 100644 public/blah/log.h create mode 100644 public/blah/math/calc.cpp create mode 100644 public/blah/math/calc.h create mode 100644 public/blah/math/circle.cpp create mode 100644 public/blah/math/circle.h create mode 100644 public/blah/math/color.cpp create mode 100644 public/blah/math/color.h create mode 100644 public/blah/math/ease.h create mode 100644 public/blah/math/line.cpp create mode 100644 public/blah/math/line.h create mode 100644 public/blah/math/mat3x2.cpp create mode 100644 public/blah/math/mat3x2.h create mode 100644 public/blah/math/mat4x4.cpp create mode 100644 public/blah/math/mat4x4.h create mode 100644 public/blah/math/point.cpp create mode 100644 public/blah/math/point.h create mode 100644 public/blah/math/quad.cpp create mode 100644 public/blah/math/quad.h create mode 100644 public/blah/math/rect.cpp create mode 100644 public/blah/math/rect.h create mode 100644 public/blah/math/rectI.cpp create mode 100644 public/blah/math/rectI.h create mode 100644 public/blah/math/stopwatch.cpp create mode 100644 public/blah/math/stopwatch.h create mode 100644 public/blah/math/vec2.cpp create mode 100644 public/blah/math/vec2.h create mode 100644 public/blah/math/vec4.h create mode 100644 public/blah/streams/bufferstream.cpp create mode 100644 public/blah/streams/bufferstream.h create mode 100644 public/blah/streams/filestream.cpp create mode 100644 public/blah/streams/filestream.h create mode 100644 public/blah/streams/memorystream.cpp create mode 100644 public/blah/streams/memorystream.h create mode 100644 public/blah/streams/stream.cpp create mode 100644 public/blah/streams/stream.h create mode 100644 public/blah/time.cpp create mode 100644 public/blah/time.h diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..8d82d20 --- /dev/null +++ b/.gitignore @@ -0,0 +1,6 @@ +build/ +out/ +temp/ +.vs/ +.vscode/ +CMakeSettings.json \ No newline at end of file diff --git a/CMakeLists.txt b/CMakeLists.txt new file mode 100644 index 0000000..77e8dd8 --- /dev/null +++ b/CMakeLists.txt @@ -0,0 +1,128 @@ +cmake_minimum_required(VERSION 3.6) +project(blah) + +# C++ version +set(CMAKE_CXX_STANDARD 17) +set(CMAKE_CXX_STANDARD_REQUIRED ON) + +add_library(blah + + public/blah.h + public/blah/app.cpp + public/blah/app.h + public/blah/filesystem.cpp + public/blah/filesystem.h + public/blah/log.cpp + public/blah/log.h + public/blah/time.cpp + public/blah/time.h + + public/blah/graphics/graphics.cpp + public/blah/graphics/graphics.h + public/blah/graphics/texture.h + public/blah/graphics/framebuffer.h + public/blah/graphics/shader.h + public/blah/graphics/mesh.h + public/blah/graphics/material.h + public/blah/graphics/material.cpp + + public/blah/input/input.cpp + public/blah/input/input.h + public/blah/input/virtual_stick.cpp + public/blah/input/virtual_stick.h + public/blah/input/virtual_button.cpp + public/blah/input/virtual_button.h + public/blah/input/virtual_axis.cpp + public/blah/input/virtual_axis.h + + public/blah/containers/list.h + public/blah/containers/stacklist.h + public/blah/containers/str.cpp + public/blah/containers/str.h + + public/blah/drawing/batch.cpp + public/blah/drawing/batch.h + public/blah/drawing/spritefont.cpp + public/blah/drawing/spritefont.h + public/blah/drawing/subtexture.cpp + public/blah/drawing/subtexture.h + + public/blah/images/aseprite.cpp + public/blah/images/aseprite.h + public/blah/images/font.cpp + public/blah/images/font.h + public/blah/images/image.cpp + public/blah/images/image.h + public/blah/images/packer.cpp + public/blah/images/packer.h + + public/blah/math/calc.cpp + public/blah/math/calc.h + public/blah/math/circle.cpp + public/blah/math/circle.h + public/blah/math/color.cpp + public/blah/math/color.h + public/blah/math/ease.h + public/blah/math/line.cpp + public/blah/math/line.h + public/blah/math/mat3x2.cpp + public/blah/math/mat3x2.h + public/blah/math/mat4x4.cpp + public/blah/math/mat4x4.h + public/blah/math/point.cpp + public/blah/math/point.h + public/blah/math/quad.h + public/blah/math/quad.cpp + public/blah/math/rect.cpp + public/blah/math/rect.h + public/blah/math/rectI.cpp + public/blah/math/rectI.h + public/blah/math/stopwatch.cpp + public/blah/math/stopwatch.h + public/blah/math/vec2.cpp + public/blah/math/vec2.h + public/blah/math/vec4.h + + public/blah/streams/bufferstream.cpp + public/blah/streams/bufferstream.h + public/blah/streams/filestream.cpp + public/blah/streams/filestream.h + public/blah/streams/memorystream.cpp + public/blah/streams/memorystream.h + public/blah/streams/stream.cpp + public/blah/streams/stream.h + + private/blah/third_party/stb_image.h + private/blah/third_party/stb_image_write.h + private/blah/third_party/stb_truetype.h + private/blah/internal/graphics.h + private/blah/internal/graphics_opengl.cpp + private/blah/internal/input.h + private/blah/internal/platform.h + private/blah/internal/platform_sdl2.cpp +) + +target_include_directories(blah + PUBLIC + $ + PRIVATE + $ +) + +# Platform Variables +set(SDL2_ENABLED true CACHE BOOL "Use SDL2 as the System implementation") +set(SDL2_INCLUDE_DIRS "" CACHE FILEPATH "SDL2 Headers") +set(SDL2_LIBRARIES "" CACHE FILEPATH "SDL2 Headers") +set(OPENGL_ENABLED true CACHE BOOL "Include OpenGL graphics implementation") + +# add OpenGL definition if we're using OpenGL +if (OPENGL_ENABLED) + add_compile_definitions(BLAH_USE_OPENGL) +endif() + +# Link and create SDL2 Definition if we're using SDL2 +if (SDL2_ENABLED) + add_compile_definitions(BLAH_USE_SDL2) + target_include_directories(blah PUBLIC "$") + target_link_libraries(blah PUBLIC ${SDL2_LIBRARIES}) +endif() \ No newline at end of file diff --git a/README.md b/README.md new file mode 100644 index 0000000..e69de29 diff --git a/private/blah/internal/graphics.h b/private/blah/internal/graphics.h new file mode 100644 index 0000000..15796d7 --- /dev/null +++ b/private/blah/internal/graphics.h @@ -0,0 +1,67 @@ +#pragma once +#include + +namespace Blah +{ + namespace Internal + { + class GraphicsDevice; + + // graphics device metadata used to instantiate and destroy + // devices of specific apis + struct GraphicsDeviceInfo + { + GfxAPI api; + bool (*supported)(); + GraphicsDevice* (*create)(); + void (*destroy)(GraphicsDevice*); + }; + + // graphics implementations + extern GraphicsDeviceInfo OpenGL_DeviceInfo; + + // graphics device + // each graphics implementation needs to implement this + // ex. one for opengl, one for vulkan, etc + class GraphicsDevice + { + public: + bool valid = false; + GraphicsInfo info; + + virtual void startup() = 0; + virtual void update() = 0; + virtual void shutdown() = 0; + virtual void before_render() = 0; + virtual void after_render() = 0; + + virtual TextureRef create_texture(int width, int height, TextureFilter filter, TextureWrap wrap_x, TextureWrap wrap_y, TextureFormat format) = 0; + virtual FrameBufferRef create_framebuffer(int width, int height, const TextureFormat* attachments, int attachmentCount) = 0; + virtual ShaderRef create_shader(const ShaderData* data) = 0; + virtual MeshRef create_mesh() = 0; + virtual void render(RenderCall* call) = 0; + virtual void clear(const FrameBufferRef& target, uint32_t rgba) = 0; + }; + + namespace Graphics + { + // Picks a graphics API based on the available APIs + GfxAPI pick_api(); + + // Initializes the Graphics with the given API + bool init(GfxAPI api); + + // Is called internally by the Application every update + void frame(); + + // Is called internally by the Application before every render + void before_render(); + + // Is called internally by the Application after every render + void after_render(); + + // Called when the Application is shutting down + void shutdown(); + } + } +} \ No newline at end of file diff --git a/private/blah/internal/graphics_opengl.cpp b/private/blah/internal/graphics_opengl.cpp new file mode 100644 index 0000000..cac2070 --- /dev/null +++ b/private/blah/internal/graphics_opengl.cpp @@ -0,0 +1,1630 @@ +#include +#include +#include +#include +#include +#include +#include +#include + +#ifdef BLAH_USE_OPENGL + +#include +#include +#include + +#ifdef _WIN32 +#include +#else +#define APIENTRY +#endif + +typedef ptrdiff_t GLintptr; +typedef ptrdiff_t GLsizeiptr; +typedef unsigned int GLenum; +typedef unsigned char GLboolean; +typedef unsigned int GLbitfield; +typedef void GLvoid; +typedef signed char GLbyte; /* 1-byte signed */ +typedef short GLshort; /* 2-byte signed */ +typedef int GLint; /* 4-byte signed */ +typedef unsigned char GLubyte; /* 1-byte unsigned */ +typedef unsigned short GLushort; /* 2-byte unsigned */ +typedef unsigned int GLuint; /* 4-byte unsigned */ +typedef int GLsizei; /* 4-byte signed */ +typedef float GLfloat; /* single precision float */ +typedef float GLclampf; /* single precision float in [0,1] */ +typedef double GLdouble; /* double precision float */ +typedef double GLclampd; /* double precision float in [0,1] */ +typedef char GLchar; + +#define GL_DONT_CARE 0x1100 +#define GL_ZERO 0x0000 +#define GL_ONE 0x0001 +#define GL_BYTE 0x1400 +#define GL_UNSIGNED_BYTE 0x1401 +#define GL_SHORT 0x1402 +#define GL_UNSIGNED_SHORT 0x1403 +#define GL_INT 0x1404 +#define GL_UNSIGNED_INT 0x1405 +#define GL_FLOAT 0x1406 +#define GL_HALF_FLOAT 0x140B +#define GL_UNSIGNED_SHORT_4_4_4_4_REV 0x8365 +#define GL_UNSIGNED_SHORT_5_5_5_1_REV 0x8366 +#define GL_UNSIGNED_INT_2_10_10_10_REV 0x8368 +#define GL_UNSIGNED_SHORT_5_6_5 0x8363 +#define GL_UNSIGNED_INT_24_8 0x84FA +#define GL_VENDOR 0x1F00 +#define GL_RENDERER 0x1F01 +#define GL_VERSION 0x1F02 +#define GL_EXTENSIONS 0x1F03 +#define GL_COLOR_BUFFER_BIT 0x4000 +#define GL_DEPTH_BUFFER_BIT 0x0100 +#define GL_STENCIL_BUFFER_BIT 0x0400 +#define GL_SCISSOR_TEST 0x0C11 +#define GL_DEPTH_TEST 0x0B71 +#define GL_STENCIL_TEST 0x0B90 +#define GL_LINE 0x1B01 +#define GL_FILL 0x1B02 +#define GL_CW 0x0900 +#define GL_CCW 0x0901 +#define GL_FRONT 0x0404 +#define GL_BACK 0x0405 +#define GL_FRONT_AND_BACK 0x0408 +#define GL_CULL_FACE 0x0B44 +#define GL_POLYGON_OFFSET_FILL 0x8037 +#define GL_TEXTURE_2D 0x0DE1 +#define GL_TEXTURE_3D 0x806F +#define GL_TEXTURE_CUBE_MAP 0x8513 +#define GL_TEXTURE_CUBE_MAP_POSITIVE_X 0x8515 +#define GL_BLEND 0x0BE2 +#define GL_SRC_COLOR 0x0300 +#define GL_ONE_MINUS_SRC_COLOR 0x0301 +#define GL_SRC_ALPHA 0x0302 +#define GL_ONE_MINUS_SRC_ALPHA 0x0303 +#define GL_DST_ALPHA 0x0304 +#define GL_ONE_MINUS_DST_ALPHA 0x0305 +#define GL_DST_COLOR 0x0306 +#define GL_ONE_MINUS_DST_COLOR 0x0307 +#define GL_SRC_ALPHA_SATURATE 0x0308 +#define GL_CONSTANT_COLOR 0x8001 +#define GL_ONE_MINUS_CONSTANT_COLOR 0x8002 +#define GL_CONSTANT_ALPHA 0x8003 +#define GL_ONE_MINUS_CONSTANT_ALPHA 0x8004 +#define GL_SRC1_ALPHA 0x8589 +#define GL_SRC1_COLOR 0x88F9 +#define GL_ONE_MINUS_SRC1_COLOR 0x88FA +#define GL_ONE_MINUS_SRC1_ALPHA 0x88FB +#define GL_MIN 0x8007 +#define GL_MAX 0x8008 +#define GL_FUNC_ADD 0x8006 +#define GL_FUNC_SUBTRACT 0x800A +#define GL_FUNC_REVERSE_SUBTRACT 0x800B +#define GL_NEVER 0x0200 +#define GL_LESS 0x0201 +#define GL_EQUAL 0x0202 +#define GL_LEQUAL 0x0203 +#define GL_GREATER 0x0204 +#define GL_NOTEQUAL 0x0205 +#define GL_GEQUAL 0x0206 +#define GL_ALWAYS 0x0207 +#define GL_INVERT 0x150A +#define GL_KEEP 0x1E00 +#define GL_REPLACE 0x1E01 +#define GL_INCR 0x1E02 +#define GL_DECR 0x1E03 +#define GL_INCR_WRAP 0x8507 +#define GL_DECR_WRAP 0x8508 +#define GL_REPEAT 0x2901 +#define GL_CLAMP_TO_EDGE 0x812F +#define GL_MIRRORED_REPEAT 0x8370 +#define GL_NEAREST 0x2600 +#define GL_LINEAR 0x2601 +#define GL_NEAREST_MIPMAP_NEAREST 0x2700 +#define GL_NEAREST_MIPMAP_LINEAR 0x2702 +#define GL_LINEAR_MIPMAP_NEAREST 0x2701 +#define GL_LINEAR_MIPMAP_LINEAR 0x2703 +#define GL_COLOR_ATTACHMENT0 0x8CE0 +#define GL_DEPTH_ATTACHMENT 0x8D00 +#define GL_STENCIL_ATTACHMENT 0x8D20 +#define GL_DEPTH_STENCIL_ATTACHMENT 0x821A +#define GL_RED 0x1903 +#define GL_RGB 0x1907 +#define GL_RGBA 0x1908 +#define GL_LUMINANCE 0x1909 +#define GL_RGB8 0x8051 +#define GL_RGBA8 0x8058 +#define GL_RGBA4 0x8056 +#define GL_RGB5_A1 0x8057 +#define GL_RGB10_A2_EXT 0x8059 +#define GL_RGBA16 0x805B +#define GL_BGRA 0x80E1 +#define GL_DEPTH_COMPONENT16 0x81A5 +#define GL_DEPTH_COMPONENT24 0x81A6 +#define GL_RG 0x8227 +#define GL_RG8 0x822B +#define GL_RG16 0x822C +#define GL_R16F 0x822D +#define GL_R32F 0x822E +#define GL_RG16F 0x822F +#define GL_RG32F 0x8230 +#define GL_RGBA32F 0x8814 +#define GL_RGBA16F 0x881A +#define GL_DEPTH24_STENCIL8 0x88F0 +#define GL_COMPRESSED_TEXTURE_FORMATS 0x86A3 +#define GL_COMPRESSED_RGBA_S3TC_DXT1_EXT 0x83F1 +#define GL_COMPRESSED_RGBA_S3TC_DXT3_EXT 0x83F2 +#define GL_COMPRESSED_RGBA_S3TC_DXT5_EXT 0x83F3 +#define GL_DEPTH_COMPONENT 0x1902 +#define GL_DEPTH_STENCIL 0x84F9 +#define GL_TEXTURE_WRAP_S 0x2802 +#define GL_TEXTURE_WRAP_T 0x2803 +#define GL_TEXTURE_WRAP_R 0x8072 +#define GL_TEXTURE_MAG_FILTER 0x2800 +#define GL_TEXTURE_MIN_FILTER 0x2801 +#define GL_TEXTURE_MAX_ANISOTROPY_EXT 0x84FE +#define GL_TEXTURE_BASE_LEVEL 0x813C +#define GL_TEXTURE_MAX_LEVEL 0x813D +#define GL_TEXTURE_LOD_BIAS 0x8501 +#define GL_PACK_ALIGNMENT 0x0D05 +#define GL_UNPACK_ALIGNMENT 0x0CF5 +#define GL_TEXTURE0 0x84C0 +#define GL_MAX_TEXTURE_IMAGE_UNITS 0x8872 +#define GL_MAX_VERTEX_TEXTURE_IMAGE_UNITS 0x8B4C +#define GL_ARRAY_BUFFER 0x8892 +#define GL_ELEMENT_ARRAY_BUFFER 0x8893 +#define GL_STREAM_DRAW 0x88E0 +#define GL_STATIC_DRAW 0x88E4 +#define GL_DYNAMIC_DRAW 0x88E8 +#define GL_MAX_VERTEX_ATTRIBS 0x8869 +#define GL_FRAMEBUFFER 0x8D40 +#define GL_READ_FRAMEBUFFER 0x8CA8 +#define GL_DRAW_FRAMEBUFFER 0x8CA9 +#define GL_RENDERBUFFER 0x8D41 +#define GL_MAX_DRAW_BUFFERS 0x8824 +#define GL_POINTS 0x0000 +#define GL_LINES 0x0001 +#define GL_LINE_STRIP 0x0003 +#define GL_TRIANGLES 0x0004 +#define GL_TRIANGLE_STRIP 0x0005 +#define GL_QUERY_RESULT 0x8866 +#define GL_QUERY_RESULT_AVAILABLE 0x8867 +#define GL_SAMPLES_PASSED 0x8914 +#define GL_MULTISAMPLE 0x809D +#define GL_MAX_SAMPLES 0x8D57 +#define GL_SAMPLE_MASK 0x8E51 +#define GL_FRAGMENT_SHADER 0x8B30 +#define GL_VERTEX_SHADER 0x8B31 +#define GL_ACTIVE_UNIFORMS 0x8B86 +#define GL_ACTIVE_ATTRIBUTES 0x8B89 +#define GL_FLOAT_VEC2 0x8B50 +#define GL_FLOAT_VEC3 0x8B51 +#define GL_FLOAT_VEC4 0x8B52 +#define GL_SAMPLER_2D 0x8B5E +#define GL_FLOAT_MAT3x2 0x8B67 +#define GL_FLOAT_MAT4 0x8B5C +#define GL_NUM_EXTENSIONS 0x821D +#define GL_DEBUG_SOURCE_API 0x8246 +#define GL_DEBUG_SOURCE_WINDOW_SYSTEM 0x8247 +#define GL_DEBUG_SOURCE_SHADER_COMPILER 0x8248 +#define GL_DEBUG_SOURCE_THIRD_PARTY 0x8249 +#define GL_DEBUG_SOURCE_APPLICATION 0x824A +#define GL_DEBUG_SOURCE_OTHER 0x824B +#define GL_DEBUG_TYPE_ERROR 0x824C +#define GL_DEBUG_TYPE_PUSH_GROUP 0x8269 +#define GL_DEBUG_TYPE_POP_GROUP 0x826A +#define GL_DEBUG_TYPE_MARKER 0x8268 +#define GL_DEBUG_TYPE_DEPRECATED_BEHAVIOR 0x824D +#define GL_DEBUG_TYPE_UNDEFINED_BEHAVIOR 0x824E +#define GL_DEBUG_TYPE_PORTABILITY 0x824F +#define GL_DEBUG_TYPE_PERFORMANCE 0x8250 +#define GL_DEBUG_TYPE_OTHER 0x8251 +#define GL_DEBUG_SEVERITY_HIGH 0x9146 +#define GL_DEBUG_SEVERITY_MEDIUM 0x9147 +#define GL_DEBUG_SEVERITY_LOW 0x9148 +#define GL_DEBUG_SEVERITY_NOTIFICATION 0x826B +#define GL_DEBUG_OUTPUT 0x92E0 +#define GL_DEBUG_OUTPUT_SYNCHRONOUS 0x8242 + +typedef void (APIENTRY *DEBUGPROC)(GLenum source, + GLenum type, + GLuint id, + GLenum severity, + GLsizei length, + const GLchar *message, + const void *userParam); + +#define GL_FUNC(name, ret, ...) \ +typedef ret (*name ## Func) (__VA_ARGS__); \ +name ## Func name + +#define GL_BIND(name) gl.name = (GL::name ## Func)(Internal::Platform::gl_get_func("gl" #name)); + +namespace Blah +{ + namespace OpenGL + { + struct GL + { + GL_FUNC(DebugMessageCallback, void, DEBUGPROC callback, const void* userParam); + GL_FUNC(GetString, const GLubyte*, GLenum name); + GL_FUNC(Flush, void, void); + GL_FUNC(Enable, void, GLenum mode); + GL_FUNC(Disable, void, GLenum mode); + GL_FUNC(Clear, void, GLenum mask); + GL_FUNC(ClearColor, void, GLfloat red, GLfloat green, GLfloat blue, GLfloat alpha); + GL_FUNC(ClearDepth, void, GLdouble depth); + GL_FUNC(ClearStencil, void, GLint stencil); + GL_FUNC(DepthMask, void, GLboolean enabled); + GL_FUNC(DepthFunc, void, GLenum func); + GL_FUNC(Viewport, void, GLint x, GLint y, GLint width, GLint height); + GL_FUNC(Scissor, void, GLint x, GLint y, GLint width, GLint height); + GL_FUNC(CullFace, void, GLenum mode); + GL_FUNC(BlendEquation, void, GLenum eq); + GL_FUNC(BlendEquationSeparate, void, GLenum modeRGB, GLenum modeAlpha); + GL_FUNC(BlendFunc, void, GLenum sfactor, GLenum dfactor); + GL_FUNC(BlendFuncSeparate, void, GLenum srcRGB, GLenum dstRGB, GLenum srcAlpha, GLenum dstAlpha); + GL_FUNC(BlendColor, void, GLfloat red, GLfloat green, GLfloat blue, GLfloat alpha); + GL_FUNC(ColorMask, void, GLboolean red, GLboolean green, GLboolean blue, GLboolean alpha); + GL_FUNC(GetIntegerv, void, GLenum name, GLint* data); + GL_FUNC(GenTextures, void, GLint n, void* textures); + GL_FUNC(GenRenderbuffers, void, GLint n, void* textures); + GL_FUNC(GenFramebuffers, void, GLint n, void* textures); + GL_FUNC(ActiveTexture, void, GLuint id); + GL_FUNC(BindTexture, void, GLenum target, GLuint id); + GL_FUNC(BindRenderbuffer, void, GLenum target, GLuint id); + GL_FUNC(BindFramebuffer, void, GLenum target, GLuint id); + GL_FUNC(TexImage2D, void, GLenum target, GLint level, GLenum internalFormat, GLint width, GLint height, GLint border, GLenum format, GLenum type, void* data); + GL_FUNC(FramebufferRenderbuffer, void, GLenum target, GLenum attachment, GLenum renderbuffertarget, GLuint renderbuffer); + GL_FUNC(FramebufferTexture2D, void, GLenum target, GLenum attachment, GLenum textarget, GLuint texture, GLint level); + GL_FUNC(TexParameteri, void, GLenum target, GLenum name, GLint param); + GL_FUNC(RenderbufferStorage, void, GLenum target, GLenum internalformat, GLint width, GLint height); + GL_FUNC(GetTexImage, void, GLenum target, GLint level, GLenum format, GLenum type, void* data); + GL_FUNC(DrawElements, void, GLenum mode, GLint count, GLenum type, void* indices); + GL_FUNC(DrawElementsInstanced, void, GLenum mode, GLint count, GLenum type, void* indices, GLint amount); + GL_FUNC(DeleteTextures, void, GLint n, GLuint* textures); + GL_FUNC(DeleteRenderbuffers, void, GLint n, GLuint* renderbuffers); + GL_FUNC(DeleteFramebuffers, void, GLint n, GLuint* textures); + GL_FUNC(GenVertexArrays, void, GLint n, GLuint* arrays); + GL_FUNC(BindVertexArray, void, GLuint id); + GL_FUNC(GenBuffers, void, GLint n, GLuint* arrays); + GL_FUNC(BindBuffer, void, GLenum target, GLuint buffer); + GL_FUNC(BufferData, void, GLenum target, GLsizeiptr size, const void* data, GLenum usage); + GL_FUNC(BufferSubData, void, GLenum target, GLintptr offset, GLsizeiptr size, const void* data); + GL_FUNC(DeleteBuffers, void, GLint n, GLuint* buffers); + GL_FUNC(DeleteVertexArrays, void, GLint n, GLuint* arrays); + GL_FUNC(EnableVertexAttribArray, void, GLuint location); + GL_FUNC(DisableVertexAttribArray, void, GLuint location); + GL_FUNC(VertexAttribPointer, void, GLuint index, GLint size, GLenum type, GLboolean normalized, GLint stride, const void* pointer); + GL_FUNC(VertexAttribDivisor, void, GLuint index, GLuint divisor); + GL_FUNC(CreateShader, GLuint, GLenum type); + GL_FUNC(AttachShader, void, GLuint program, GLuint shader); + GL_FUNC(DetachShader, void, GLuint program, GLuint shader); + GL_FUNC(DeleteShader, void, GLuint shader); + GL_FUNC(ShaderSource, void, GLuint shader, GLsizei count, const GLchar** string, const GLint* length); + GL_FUNC(CompileShader, void, GLuint shader); + GL_FUNC(GetShaderiv, void, GLuint shader, GLenum pname, GLint* result); + GL_FUNC(GetShaderInfoLog, void, GLuint shader, GLint maxLength, GLsizei* length, GLchar* infoLog); + GL_FUNC(CreateProgram, GLuint, ); + GL_FUNC(DeleteProgram, void, GLuint program); + GL_FUNC(LinkProgram, void, GLuint program); + GL_FUNC(GetProgramiv, void, GLuint program, GLenum pname, GLint* result); + GL_FUNC(GetProgramInfoLog, void, GLuint program, GLint maxLength, GLsizei* length, GLchar* infoLog); + GL_FUNC(GetActiveUniform, void, GLuint program, GLuint index, GLint bufSize, GLsizei* length, GLint* size, GLenum* type, GLchar* name); + GL_FUNC(GetActiveAttrib, void, GLuint program, GLuint index, GLint bufSize, GLsizei* length, GLint* size, GLenum* type, GLchar* name); + GL_FUNC(UseProgram, void, GLuint program); + GL_FUNC(GetUniformLocation, GLint, GLuint program, const GLchar* name); + GL_FUNC(GetAttribLocation, GLint, GLuint program, const GLchar* name); + GL_FUNC(Uniform1f, void, GLint location, GLfloat v0); + GL_FUNC(Uniform2f, void, GLint location, GLfloat v0, GLfloat v1); + GL_FUNC(Uniform3f, void, GLint location, GLfloat v0, GLfloat v1, GLfloat v2); + GL_FUNC(Uniform4f, void, GLint location, GLfloat v0, GLfloat v1, GLfloat v2, GLfloat v3); + GL_FUNC(Uniform1fv, void, GLint location, GLint count, const GLfloat* value); + GL_FUNC(Uniform2fv, void, GLint location, GLint count, const GLfloat* value); + GL_FUNC(Uniform3fv, void, GLint location, GLint count, const GLfloat* value); + GL_FUNC(Uniform4fv, void, GLint location, GLint count, const GLfloat* value); + GL_FUNC(Uniform1i, void, GLint location, GLint v0); + GL_FUNC(Uniform2i, void, GLint location, GLint v0, GLint v1); + GL_FUNC(Uniform3i, void, GLint location, GLint v0, GLint v1, GLint v2); + GL_FUNC(Uniform4i, void, GLint location, GLint v0, GLint v1, GLint v2, GLint v3); + GL_FUNC(Uniform1iv, void, GLint location, GLint count, const GLint* value); + GL_FUNC(Uniform2iv, void, GLint location, GLint count, const GLint* value); + GL_FUNC(Uniform3iv, void, GLint location, GLint count, const GLint* value); + GL_FUNC(Uniform4iv, void, GLint location, GLint count, const GLint* value); + GL_FUNC(Uniform1ui, void, GLint location, GLuint v0); + GL_FUNC(Uniform2ui, void, GLint location, GLuint v0, GLuint v1); + GL_FUNC(Uniform3ui, void, GLint location, GLuint v0, GLuint v1, GLuint v2); + GL_FUNC(Uniform4ui, void, GLint location, GLuint v0, GLuint v1, GLuint v2, GLuint v3); + GL_FUNC(Uniform1uiv, void, GLint location, GLint count, const GLint* value); + GL_FUNC(Uniform2uiv, void, GLint location, GLint count, const GLint* value); + GL_FUNC(Uniform3uiv, void, GLint location, GLint count, const GLint* value); + GL_FUNC(Uniform4uiv, void, GLint location, GLint count, const GLint* value); + GL_FUNC(UniformMatrix2fv, void, GLint location, GLint count, GLboolean transpose, const GLfloat* value); + GL_FUNC(UniformMatrix3fv, void, GLint location, GLint count, GLboolean transpose, const GLfloat* value); + GL_FUNC(UniformMatrix4fv, void, GLint location, GLint count, GLboolean transpose, const GLfloat* value); + GL_FUNC(UniformMatrix2x3fv, void, GLint location, GLint count, GLboolean transpose, const GLfloat* value); + GL_FUNC(UniformMatrix3x2fv, void, GLint location, GLint count, GLboolean transpose, const GLfloat* value); + GL_FUNC(UniformMatrix2x4fv, void, GLint location, GLint count, GLboolean transpose, const GLfloat* value); + GL_FUNC(UniformMatrix4x2fv, void, GLint location, GLint count, GLboolean transpose, const GLfloat* value); + GL_FUNC(UniformMatrix3x4fv, void, GLint location, GLint count, GLboolean transpose, const GLfloat* value); + GL_FUNC(UniformMatrix4x3fv, void, GLint location, GLint count, GLboolean transpose, const GLfloat* value); + GL_FUNC(PixelStorei, void, GLenum pname, GLint param); + }; + + // static state + // TODO: + // this should probably be part of the Device implementation ... + void* context; + GL gl; + GLint maxColorAttachments; + GLint maxElementIndices; + GLint maxElementVertices; + GLint maxRenderBufferSize; + GLint maxSamples; + GLint maxTextureImageUnits; + GLint maxTextureSize; + + void APIENTRY gl_message_callback(GLenum source, GLenum type, GLuint id, GLenum severity, GLsizei length, const GLchar* message, const void* userParam) + { + // these are basically never useful + if (severity == GL_DEBUG_SEVERITY_NOTIFICATION && + type == GL_DEBUG_TYPE_OTHER) + return; + + const char* typeName; + const char* severityName; + + switch (type) + { + case GL_DEBUG_TYPE_ERROR: typeName = "ERROR"; break; + case GL_DEBUG_TYPE_DEPRECATED_BEHAVIOR: typeName = "DEPRECATED BEHAVIOR"; break; + case GL_DEBUG_TYPE_MARKER: typeName = "MARKER"; break; + case GL_DEBUG_TYPE_OTHER: typeName = "OTHER"; break; + case GL_DEBUG_TYPE_PERFORMANCE: typeName = "PEROFRMANCE"; break; + case GL_DEBUG_TYPE_POP_GROUP: typeName = "POP GROUP"; break; + case GL_DEBUG_TYPE_PORTABILITY: typeName = "PORTABILITY"; break; + case GL_DEBUG_TYPE_PUSH_GROUP: typeName = "PUSH GROUP"; break; + case GL_DEBUG_TYPE_UNDEFINED_BEHAVIOR: typeName = "UNDEFINED BEHAVIOR"; break; + } + + switch (severity) + { + case GL_DEBUG_SEVERITY_HIGH: severityName = "HIGH"; break; + case GL_DEBUG_SEVERITY_MEDIUM: severityName = "MEDIUM"; break; + case GL_DEBUG_SEVERITY_LOW: severityName = "LOW"; break; + case GL_DEBUG_SEVERITY_NOTIFICATION: severityName = "NOTIFICATION"; break; + } + + if (type == GL_DEBUG_TYPE_ERROR) + Log::error("GL (%s:%s) %s", typeName, severityName, message); + else if (severity != GL_DEBUG_SEVERITY_NOTIFICATION) + Log::warn("GL (%s:%s) %s", typeName, severityName, message); + else + Log::print("GL (%s) %s", typeName, message); + } + + GLuint gl_mesh_assign_attributes(GLuint buffer, GLenum buffer_type, const VertexAttribute* vertex_attributes, int vertex_attribute_count, int stride, GLint divisor) + { + // bind + gl.BindBuffer(buffer_type, buffer); + + // TODO: disable existing enabled attributes .. + // ... + // ... + + // enable attributes + size_t ptr = 0; + for (int n = 0; n < vertex_attribute_count; n++) + { + const VertexAttribute* attrib = (vertex_attributes + n); + + for (int i = 0, loc = 0; i < (int)attrib->components; i += 4, loc++) + { + int components = attrib->components - i; + if (components > 4) + components = 4; + + GLenum type = GL_UNSIGNED_BYTE; + size_t component_size = 0; + if (attrib->type == VertexAttributeType::Byte) + { + type = GL_UNSIGNED_BYTE; + component_size = 1; + } + else if (attrib->type == VertexAttributeType::Short) + { + type = GL_SHORT; + component_size = 2; + } + else if (attrib->type == VertexAttributeType::Int) + { + type = GL_INT; + component_size = 4; + } + else if (attrib->type == VertexAttributeType::Float) + { + type = GL_FLOAT; + component_size = 4; + } + + uint32_t location = (uint32_t)(attrib->index + loc); + gl.EnableVertexAttribArray(location); + gl.VertexAttribPointer(location, components, type, attrib->normalized, stride, (void*)ptr); + gl.VertexAttribDivisor(location, divisor); + + ptr += components * component_size; + } + } + + return stride; + } + + GLenum gl_get_blend_func(BlendOp operation) + { + switch (operation) + { + case BlendOp::Add: return GL_FUNC_ADD; + case BlendOp::Subtract: return GL_FUNC_SUBTRACT; + case BlendOp::ReverseSubtract: return GL_FUNC_REVERSE_SUBTRACT; + case BlendOp::Min: return GL_MIN; + case BlendOp::Max: return GL_MAX; + }; + return GL_FUNC_ADD; + } + + GLenum gl_get_blend_factor(BlendFactor factor) + { + switch (factor) + { + case BlendFactor::Zero: return GL_ZERO; + case BlendFactor::One: return GL_ONE; + case BlendFactor::SrcColor: return GL_SRC_COLOR; + case BlendFactor::OneMinusSrcColor: return GL_ONE_MINUS_SRC_COLOR; + case BlendFactor::DstColor: return GL_DST_COLOR; + case BlendFactor::OneMinusDstColor: return GL_ONE_MINUS_DST_COLOR; + case BlendFactor::SrcAlpha: return GL_SRC_ALPHA; + case BlendFactor::OneMinusSrcAlpha: return GL_ONE_MINUS_SRC_ALPHA; + case BlendFactor::DstAlpha: return GL_DST_ALPHA; + case BlendFactor::OneMinusDstAlpha: return GL_ONE_MINUS_DST_ALPHA; + case BlendFactor::ConstantColor: return GL_CONSTANT_COLOR; + case BlendFactor::OneMinusConstantColor: return GL_ONE_MINUS_CONSTANT_COLOR; + case BlendFactor::ConstantAlpha: return GL_CONSTANT_ALPHA; + case BlendFactor::OneMinusConstantAlpha: return GL_ONE_MINUS_CONSTANT_ALPHA; + case BlendFactor::SrcAlphaSaturate: return GL_SRC_ALPHA_SATURATE; + case BlendFactor::Src1Color: return GL_SRC1_COLOR; + case BlendFactor::OneMinusSrc1Color: return GL_ONE_MINUS_SRC1_COLOR; + case BlendFactor::Src1Alpha: return GL_SRC1_ALPHA; + case BlendFactor::OneMinusSrc1Alpha: return GL_ONE_MINUS_SRC1_ALPHA; + }; + + return GL_ZERO; + } + + class OpenGL_Texture : public Texture + { + private: + GLuint m_id; + int m_width; + int m_height; + TextureWrap m_wrap_x; + TextureWrap m_wrap_y; + TextureFilter m_filter; + TextureFormat m_format; + GLenum m_gl_internal_format; + GLenum m_gl_format; + GLenum m_gl_type; + + public: + bool framebuffer_parent; + + OpenGL_Texture(int width, int height, TextureFilter filter, TextureWrap wrap_x, TextureWrap wrap_y, TextureFormat format) + { + m_id = 0; + m_width = width; + m_height = height; + m_wrap_x = wrap_x; + m_wrap_y = wrap_y; + m_filter = filter; + m_format = format; + framebuffer_parent = false; + + if (width > maxTextureSize || height > maxTextureSize) + { + Log::error("Exceeded Max Texture Size of %i", maxTextureSize); + return; + } + + if (format == TextureFormat::R) + { + m_gl_internal_format = GL_RED; + m_gl_format = GL_RED; + m_gl_type = GL_UNSIGNED_BYTE; + } + else if (format == TextureFormat::RG) + { + m_gl_internal_format = GL_RG; + m_gl_format = GL_RG; + m_gl_type = GL_UNSIGNED_BYTE; + } + else if (format == TextureFormat::RGB) + { + m_gl_internal_format = GL_RGB; + m_gl_format = GL_RGB; + m_gl_type = GL_UNSIGNED_BYTE; + } + else if (format == TextureFormat::RGBA) + { + m_gl_internal_format = GL_RGBA; + m_gl_format = GL_RGBA; + m_gl_type = GL_UNSIGNED_BYTE; + } + else if (format == TextureFormat::DepthStencil) + { + m_gl_internal_format = GL_DEPTH24_STENCIL8; + m_gl_format = GL_DEPTH_STENCIL; + m_gl_type = GL_UNSIGNED_INT_24_8; + } + else + { + Log::error("Invalid Texture Format %i", format); + return; + } + + gl.GenTextures(1, &m_id); + gl.ActiveTexture(GL_TEXTURE0); + gl.BindTexture(GL_TEXTURE_2D, m_id); + + gl.TexImage2D(GL_TEXTURE_2D, 0, m_gl_internal_format, width, height, 0, m_gl_format, m_gl_type, nullptr); + gl.TexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, (m_filter == TextureFilter::Nearest ? GL_NEAREST : GL_LINEAR)); + gl.TexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, (m_filter == TextureFilter::Nearest ? GL_NEAREST : GL_LINEAR)); + gl.TexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, (m_wrap_x == TextureWrap::Clamp ? GL_CLAMP_TO_EDGE : GL_REPEAT)); + gl.TexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, (m_wrap_y == TextureWrap::Clamp ? GL_CLAMP_TO_EDGE : GL_REPEAT)); + } + + ~OpenGL_Texture() + { + if (m_id > 0) + gl.DeleteTextures(1, &m_id); + } + + GLuint gl_id() const + { + return m_id; + } + + virtual int width() const override + { + return m_width; + } + + virtual int height() const override + { + return m_height; + } + + virtual TextureFormat format() const override + { + return m_format; + } + + virtual void set_filter(TextureFilter filter) override + { + m_filter = filter; + + gl.BindTexture(GL_TEXTURE_2D, m_id); + gl.TexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, (filter == TextureFilter::Nearest ? GL_NEAREST : GL_LINEAR)); + gl.TexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, (filter == TextureFilter::Nearest ? GL_NEAREST : GL_LINEAR)); + } + + virtual TextureFilter get_filter() const override + { + return m_filter; + } + + virtual void set_wrap(TextureWrap x, TextureWrap y) override + { + m_wrap_x = x; + m_wrap_y = y; + + gl.BindTexture(GL_TEXTURE_2D, m_id); + gl.TexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, (x == TextureWrap::Clamp ? GL_CLAMP_TO_EDGE : GL_REPEAT)); + gl.TexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, (y == TextureWrap::Clamp ? GL_CLAMP_TO_EDGE : GL_REPEAT)); + } + + virtual TextureWrap get_wrap_x() const override + { + return m_wrap_x; + } + + virtual TextureWrap get_wrap_y() const override + { + return m_wrap_y; + } + + virtual void set_data(unsigned char* data) override + { + gl.ActiveTexture(GL_TEXTURE0); + gl.BindTexture(GL_TEXTURE_2D, m_id); + gl.TexImage2D(GL_TEXTURE_2D, 0, m_gl_internal_format, m_width, m_height, 0, m_gl_format, m_gl_type, data); + } + + virtual void get_data(unsigned char* data) override + { + gl.ActiveTexture(GL_TEXTURE0); + gl.BindTexture(GL_TEXTURE_2D, m_id); + gl.GetTexImage(GL_TEXTURE_2D, 0, m_gl_internal_format, m_gl_type, data); + } + + virtual bool is_framebuffer() const override + { + return framebuffer_parent; + } + + virtual bool is_valid() const override + { + return m_id > 0; + } + + virtual void dispose() override + { + if (m_id > 0) + { + gl.DeleteTextures(1, &m_id); + m_id = 0; + } + } + + }; + + class OpenGL_FrameBuffer : public FrameBuffer + { + private: + GLuint m_id; + int m_width; + int m_height; + StackList m_attachments; + + public: + + OpenGL_FrameBuffer(int width, int height, const TextureFormat* attachments, int attachmentCount) + { + gl.GenFramebuffers(1, &m_id); + m_width = width; + m_height = height; + + gl.BindFramebuffer(GL_FRAMEBUFFER, m_id); + + for (int i = 0; i < attachmentCount; i++) + { + auto tex = Graphics::create_texture(width, height, attachments[i]); + auto gltex = ((OpenGL_Texture*)tex.get()); + + gltex->framebuffer_parent = true; + m_attachments.add(tex); + + if (attachments[i] != TextureFormat::DepthStencil) + { + gl.FramebufferTexture2D(GL_FRAMEBUFFER, GL_COLOR_ATTACHMENT0 + i, GL_TEXTURE_2D, gltex->gl_id(), 0); + } + else + { + gl.FramebufferTexture2D(GL_FRAMEBUFFER, GL_DEPTH_STENCIL_ATTACHMENT, GL_TEXTURE_2D, gltex->gl_id(), 0); + } + } + } + + ~OpenGL_FrameBuffer() + { + if (m_id > 0) + { + gl.DeleteFramebuffers(1, &m_id); + m_id = 0; + } + } + + GLuint gl_id() const + { + return m_id; + } + + virtual Attachments& attachments() override + { + return m_attachments; + } + + virtual const Attachments& attachments() const override + { + return m_attachments; + } + + virtual int width() const override + { + return m_width; + } + + virtual int height() const override + { + return m_height; + } + + virtual bool is_valid() const override + { + return m_id > 0; + } + + virtual void dispose() override + { + if (m_id > 0) + { + m_attachments.clear(); + gl.DeleteFramebuffers(1, &m_id); + m_id = 0; + } + } + }; + + class OpenGL_Shader : public Shader + { + private: + GLuint m_id; + Attributes m_attributes; + Uniforms m_uniforms; + + public: + GLint uniforms_loc[BLAH_UNIFORMS]; + + OpenGL_Shader(const ShaderData* data) + { + m_id = 0; + + // vertex shader + if (data->vertex == nullptr) + { + Log::error("Vertex Shader is required"); + return; + } + + if (data->fragment == nullptr) + { + Log::error("Fragment Shader is required"); + return; + } + + GLchar log[1024]; + GLsizei log_length = 0; + + GLuint vertex_shader = gl.CreateShader(GL_VERTEX_SHADER); + { + const GLchar* source = (const GLchar*)data->vertex; + gl.ShaderSource(vertex_shader, 1, &source, nullptr); + gl.CompileShader(vertex_shader); + gl.GetShaderInfoLog(vertex_shader, 1024, &log_length, log); + + if (log_length > 0) + { + gl.DeleteShader(vertex_shader); + Log::error(log); + return; + } + } + + GLuint fragment_shader = gl.CreateShader(GL_FRAGMENT_SHADER); + { + const GLchar* source = (const GLchar*)data->fragment; + gl.ShaderSource(fragment_shader, 1, &source, nullptr); + gl.CompileShader(fragment_shader); + gl.GetShaderInfoLog(fragment_shader, 1024, &log_length, log); + + if (log_length > 0) + { + gl.DeleteShader(fragment_shader); + Log::error(log); + return; + } + } + + GLuint id = gl.CreateProgram(); + gl.AttachShader(id, vertex_shader); + gl.AttachShader(id, fragment_shader); + gl.LinkProgram(id); + gl.GetProgramInfoLog(id, 1024, &log_length, log); + + if (log_length > 0) + { + gl.DetachShader(id, vertex_shader); + gl.DetachShader(id, fragment_shader); + gl.DeleteShader(vertex_shader); + gl.DeleteShader(fragment_shader); + + Log::error(log); + return; + } + + // ready to go + m_id = id; + + // get attributes + { + GLint active_attributes = 0; + gl.GetProgramiv(id, GL_ACTIVE_ATTRIBUTES, &active_attributes); + if (active_attributes > BLAH_ATTRIBUTES) + { + Log::warn("Exceeding maximum shader attributes (%i / %i)", active_attributes, BLAH_ATTRIBUTES); + active_attributes = BLAH_ATTRIBUTES; + } + + for (int i = 0; i < active_attributes; i++) + { + GLsizei length; + GLsizei size; + GLenum type; + GLchar name[BLAH_ATTRIBUTE_NAME]; + + gl.GetActiveAttrib(id, i, BLAH_ATTRIBUTE_NAME - 1, &length, &size, &type, name); + name[length] = '\0'; + + auto attr = m_attributes.expand(); + attr->name.append(name); + // TOOD: set these? + attr->semantic_name = ""; + attr->semantic_location = 0; + } + } + + // get uniforms + { + GLint active_uniforms = 0; + gl.GetProgramiv(id, GL_ACTIVE_UNIFORMS, &active_uniforms); + if (active_uniforms > BLAH_UNIFORMS) + { + Log::warn("Exceeding maximum shader uniforms (%i / %i)", active_uniforms, BLAH_ATTRIBUTES); + active_uniforms = BLAH_UNIFORMS; + } + + for (int i = 0; i < active_uniforms; i++) + { + GLsizei length; + GLsizei size; + GLenum type; + GLchar name[BLAH_UNIFORM_NAME]; + + gl.GetActiveUniform(id, i, BLAH_UNIFORM_NAME - 1, &length, &size, &type, name); + name[length] = '\0'; + + // array names end with "[0]", and we don't want that + for (int n = 0; n < BLAH_UNIFORM_NAME; n++) + if (name[n] == '[') + { + if (name[n + 1] == '0' && name[n + 2] == ']') + { + name[n] = '\0'; + break; + } + } + + auto uniform = m_uniforms.expand(); + uniform->name = name; + uniform->type = UniformType::None; + uniform->array_length = size; + uniforms_loc[i] = gl.GetUniformLocation(id, name); + + if (type == GL_FLOAT) + uniform->type = UniformType::Float; + else if (type == GL_FLOAT_VEC2) + uniform->type = UniformType::Float2; + else if (type == GL_FLOAT_VEC3) + uniform->type = UniformType::Float3; + else if (type == GL_FLOAT_VEC4) + uniform->type = UniformType::Float4; + else if (type == GL_FLOAT_MAT3x2) + uniform->type = UniformType::Mat3x2; + else if (type == GL_FLOAT_MAT4) + uniform->type = UniformType::Mat4x4; + else if (type == GL_SAMPLER_2D) + uniform->type = UniformType::Texture; + else + { + Log::error("Unsupported Uniform Type. Must be either FLOAT, MAT3x2, MAT4, or SAMPLER_2D"); + break; + } + } + } + } + + ~OpenGL_Shader() + { + gl.DeleteProgram(m_id); + m_id = 0; + } + + GLuint gl_id() const + { + return m_id; + } + + virtual Uniforms& uniforms() override + { + return m_uniforms; + } + + virtual const Uniforms& uniforms() const override + { + return m_uniforms; + } + + virtual Attributes& attributes() override + { + return m_attributes; + } + + virtual const Attributes& attributes() const override + { + return m_attributes; + } + + virtual bool is_valid() const override + { + return m_id > 0; + } + + virtual void dispose() override + { + gl.DeleteProgram(m_id); + m_id = 0; + } + }; + + class OpenGL_Mesh : public Mesh + { + private: + GLuint m_id; + GLuint m_index_buffer; + GLuint m_vertex_buffer; + GLuint m_instance_buffer; + int64_t m_index_count; + int64_t m_vertex_count; + int64_t m_instance_count; + uint16_t m_vertex_size; + uint16_t m_instance_size; + uint8_t m_vertex_attribs_enabled; + uint8_t m_instance_attribs_enabled; + GLuint m_vertex_attribs[BLAH_ATTRIBUTES]; + GLuint m_instance_attribs[BLAH_ATTRIBUTES]; + + public: + + OpenGL_Mesh() + { + m_id = 0; + m_index_buffer = 0; + m_vertex_buffer = 0; + m_instance_buffer = 0; + m_index_count = 0; + m_vertex_count = 0; + m_instance_count = 0; + m_vertex_size = 0; + m_instance_size = 0; + m_vertex_attribs_enabled = 0; + m_instance_attribs_enabled = 0; + + gl.GenVertexArrays(1, &m_id); + } + + ~OpenGL_Mesh() + { + if (m_vertex_buffer != 0) + gl.DeleteBuffers(1, &m_vertex_buffer); + if (m_index_buffer != 0) + gl.DeleteBuffers(1, &m_index_buffer); + if (m_instance_buffer != 0) + gl.DeleteBuffers(1, &m_instance_buffer); + if (m_id != 0) + gl.DeleteVertexArrays(1, &m_id); + m_id = 0; + } + + GLuint gl_id() const + { + return m_id; + } + + virtual void vertex_format(const VertexAttribute* attributes, int attribute_count, int stride = -1) override + { + gl.BindVertexArray(m_id); + { + if (m_vertex_buffer == 0) + gl.GenBuffers(1, &(m_vertex_buffer)); + + m_vertex_size = gl_mesh_assign_attributes(m_vertex_buffer, GL_ARRAY_BUFFER, attributes, attribute_count, stride, 0); + } + gl.BindVertexArray(0); + } + + virtual void instance_format(const VertexAttribute* attributes, int attribute_count, int stride = -1) override + { + gl.BindVertexArray(m_id); + { + if (m_instance_buffer == 0) + gl.GenBuffers(1, &(m_instance_buffer)); + + m_instance_size = gl_mesh_assign_attributes(m_instance_buffer, GL_ARRAY_BUFFER, attributes, attribute_count, stride, 1); + } + gl.BindVertexArray(0); + } + + virtual void index_data(const void* indices, int64_t count) override + { + m_index_count = count; + + gl.BindVertexArray(m_id); + { + if (m_index_buffer == 0) + gl.GenBuffers(1, &(m_index_buffer)); + + gl.BindBuffer(GL_ELEMENT_ARRAY_BUFFER, m_index_buffer); + gl.BufferData(GL_ELEMENT_ARRAY_BUFFER, sizeof(int) * count, indices, GL_DYNAMIC_DRAW); + } + gl.BindVertexArray(0); + } + + virtual void vertex_data(const void* vertices, int64_t count) override + { + if (m_vertex_buffer == 0 || m_vertex_size <= 0) + { + Log::error("You must assign a Vertex Format before setting Vertex Data"); + } + else + { + m_vertex_count = count; + + gl.BindVertexArray(m_id); + { + gl.BindBuffer(GL_ARRAY_BUFFER, m_vertex_buffer); + gl.BufferData(GL_ARRAY_BUFFER, m_vertex_size * count, vertices, GL_DYNAMIC_DRAW); + } + gl.BindVertexArray(0); + } + } + + virtual void instance_data(const void* instances, int64_t count) override + { + if (m_instance_buffer == 0 || m_instance_size <= 0) + { + Log::error("You must assign an Instance Format before setting Instance Data"); + } + else + { + m_instance_count = count; + + gl.BindVertexArray(m_id); + { + gl.BindBuffer(GL_ARRAY_BUFFER, m_instance_buffer); + gl.BufferData(GL_ARRAY_BUFFER, m_instance_size * count, instances, GL_DYNAMIC_DRAW); + } + gl.BindVertexArray(0); + } + } + + virtual int64_t index_count() const override + { + return m_index_count; + } + + virtual int64_t vertex_count() const override + { + return m_vertex_count; + } + + virtual int64_t instance_count() const override + { + return m_instance_count; + } + + virtual bool is_valid() const override + { + return m_id > 0; + } + + virtual void dispose() override + { + if (m_vertex_buffer != 0) + gl.DeleteBuffers(1, &m_vertex_buffer); + if (m_index_buffer != 0) + gl.DeleteBuffers(1, &m_index_buffer); + if (m_instance_buffer != 0) + gl.DeleteBuffers(1, &m_instance_buffer); + if (m_id != 0) + gl.DeleteVertexArrays(1, &m_id); + m_id = 0; + } + }; + + class Device : public Internal::GraphicsDevice + { + virtual void startup() override + { + valid = true; + + // create gl context + context = Internal::Platform::gl_context_create(); + if (context == nullptr) + { + Log::error("Failed to create OpenGL Context"); + valid = false; + return; + } + Internal::Platform::gl_context_make_current(context); + + // bind opengl functions + GL_BIND(DebugMessageCallback); + GL_BIND(GetString); + GL_BIND(Flush); + GL_BIND(Enable); + GL_BIND(Disable); + GL_BIND(Clear); + GL_BIND(ClearColor); + GL_BIND(ClearDepth); + GL_BIND(ClearStencil); + GL_BIND(DepthMask); + GL_BIND(DepthFunc); + GL_BIND(Viewport); + GL_BIND(Scissor); + GL_BIND(CullFace); + GL_BIND(BlendEquation); + GL_BIND(BlendEquationSeparate); + GL_BIND(BlendFunc); + GL_BIND(BlendFuncSeparate); + GL_BIND(BlendColor); + GL_BIND(ColorMask); + GL_BIND(GetIntegerv); + GL_BIND(GenTextures); + GL_BIND(GenRenderbuffers); + GL_BIND(GenFramebuffers); + GL_BIND(ActiveTexture); + GL_BIND(BindTexture); + GL_BIND(BindRenderbuffer); + GL_BIND(BindFramebuffer); + GL_BIND(TexImage2D); + GL_BIND(FramebufferRenderbuffer); + GL_BIND(FramebufferTexture2D); + GL_BIND(TexParameteri); + GL_BIND(RenderbufferStorage); + GL_BIND(GetTexImage); + GL_BIND(DrawElements); + GL_BIND(DrawElementsInstanced); + GL_BIND(DeleteTextures); + GL_BIND(DeleteRenderbuffers); + GL_BIND(DeleteFramebuffers); + GL_BIND(GenVertexArrays); + GL_BIND(BindVertexArray); + GL_BIND(GenBuffers); + GL_BIND(BindBuffer); + GL_BIND(BufferData); + GL_BIND(BufferSubData); + GL_BIND(DeleteBuffers); + GL_BIND(DeleteVertexArrays); + GL_BIND(EnableVertexAttribArray); + GL_BIND(DisableVertexAttribArray); + GL_BIND(VertexAttribPointer); + GL_BIND(VertexAttribDivisor); + GL_BIND(CreateShader); + GL_BIND(AttachShader); + GL_BIND(DetachShader); + GL_BIND(DeleteShader); + GL_BIND(ShaderSource); + GL_BIND(CompileShader); + GL_BIND(GetShaderiv); + GL_BIND(GetShaderInfoLog); + GL_BIND(CreateProgram); + GL_BIND(DeleteProgram); + GL_BIND(LinkProgram); + GL_BIND(GetProgramiv); + GL_BIND(GetProgramInfoLog); + GL_BIND(GetActiveUniform); + GL_BIND(GetActiveAttrib); + GL_BIND(UseProgram); + GL_BIND(GetUniformLocation); + GL_BIND(GetAttribLocation); + GL_BIND(Uniform1f); + GL_BIND(Uniform2f); + GL_BIND(Uniform3f); + GL_BIND(Uniform4f); + GL_BIND(Uniform1fv); + GL_BIND(Uniform2fv); + GL_BIND(Uniform3fv); + GL_BIND(Uniform4fv); + GL_BIND(Uniform1i); + GL_BIND(Uniform2i); + GL_BIND(Uniform3i); + GL_BIND(Uniform4i); + GL_BIND(Uniform1iv); + GL_BIND(Uniform2iv); + GL_BIND(Uniform3iv); + GL_BIND(Uniform4iv); + GL_BIND(Uniform1ui); + GL_BIND(Uniform2ui); + GL_BIND(Uniform3ui); + GL_BIND(Uniform4ui); + GL_BIND(Uniform1uiv); + GL_BIND(Uniform2uiv); + GL_BIND(Uniform3uiv); + GL_BIND(Uniform4uiv); + GL_BIND(UniformMatrix2fv); + GL_BIND(UniformMatrix3fv); + GL_BIND(UniformMatrix4fv); + GL_BIND(UniformMatrix2x3fv); + GL_BIND(UniformMatrix3x2fv); + GL_BIND(UniformMatrix2x4fv); + GL_BIND(UniformMatrix4x2fv); + GL_BIND(UniformMatrix3x4fv); + GL_BIND(UniformMatrix4x3fv); + GL_BIND(PixelStorei); + + // bind debug message callback + if (gl.DebugMessageCallback != nullptr) + { + gl.Enable(GL_DEBUG_OUTPUT); + gl.Enable(GL_DEBUG_OUTPUT_SYNCHRONOUS); + gl.DebugMessageCallback(gl_message_callback, nullptr); + } + + // get opengl info + gl.GetIntegerv(0x8CDF, &maxColorAttachments); + gl.GetIntegerv(0x80E9, &maxElementIndices); + gl.GetIntegerv(0x80E8, &maxElementVertices); + gl.GetIntegerv(0x84E8, &maxRenderBufferSize); + gl.GetIntegerv(0x8D57, &maxSamples); + gl.GetIntegerv(0x8872, &maxTextureImageUnits); + gl.GetIntegerv(0x0D33, &maxTextureSize); + + // log + Log::print("OpenGL %s, %s", + gl.GetString(GL_VERSION), + gl.GetString(GL_RENDERER)); + + // don't include row padding + gl.PixelStorei(GL_PACK_ALIGNMENT, 1); + gl.PixelStorei(GL_UNPACK_ALIGNMENT, 1); + + // assign info + info.api = GfxAPI::OpenGL; + info.instancing = true; + info.origin_bottom_left = true; + info.max_texture_size = maxTextureSize; + } + + virtual void shutdown() override + { + Internal::Platform::gl_context_destroy(context); + context = nullptr; + } + + virtual void update() override {} + virtual void before_render() override {} + virtual void after_render() override {} + + virtual TextureRef create_texture(int width, int height, TextureFilter filter, TextureWrap wrap_x, TextureWrap wrap_y, TextureFormat format) override + { + return TextureRef(new OpenGL_Texture(width, height, filter, wrap_x, wrap_y, format)); + } + + virtual FrameBufferRef create_framebuffer(int width, int height, const TextureFormat* attachments, int attachmentCount) override + { + return FrameBufferRef(new OpenGL_FrameBuffer(width, height, attachments, attachmentCount)); + } + + virtual ShaderRef create_shader(const ShaderData* data) override + { + return ShaderRef(new OpenGL_Shader(data)); + } + + virtual MeshRef create_mesh() override + { + return MeshRef(new OpenGL_Mesh()); + } + + virtual void render(RenderCall* call) override + { + // Bind the Target + Point size; + if (!call->target || !call->target->is_valid()) + { + gl.BindFramebuffer(GL_FRAMEBUFFER, 0); + size.x = App::draw_width(); + size.y = App::draw_height(); + } + else + { + gl.BindFramebuffer(GL_FRAMEBUFFER, ((OpenGL_FrameBuffer*)call->target.get())->gl_id()); + size.x = call->target->width(); + size.y = call->target->height(); + } + + // Use the Shader + // TODO: I don't love how material values are assigned or set here + { + OpenGL_Shader* shader = (OpenGL_Shader*)(call->material->shader().get()); + gl.UseProgram(shader->gl_id()); + + // upload uniform values + int texture_slot = 0; + GLint texture_ids[64]; + + auto& uniforms = shader->uniforms(); + for (int i = 0; i < uniforms.count(); i++) + { + auto location = shader->uniforms_loc[i]; + auto& uniform = uniforms[i]; + + // Sampler 2D + if (uniform.type == UniformType::Texture) + { + for (int n = 0; n < uniform.array_length; n++) + { + auto tex = call->material->get_texture(i, n); + + gl.ActiveTexture(GL_TEXTURE0 + texture_slot); + + if (!tex) + { + gl.BindTexture(GL_TEXTURE_2D, 0); + } + else + { + gl.BindTexture(GL_TEXTURE_2D, ((OpenGL_Texture*)tex.get())->gl_id()); + } + + texture_ids[n] = texture_slot; + texture_slot++; + } + + gl.Uniform1iv(location, (GLint)uniform.array_length, &texture_ids[0]); + } + // Float + else if (uniform.type == UniformType::Float) + { + gl.Uniform1fv(location, (GLint)uniform.array_length, (const GLfloat*)call->material->get_value(i)); + } + // Float2 + else if (uniform.type == UniformType::Float2) + { + gl.Uniform2fv(location, (GLint)uniform.array_length, (const GLfloat*)call->material->get_value(i)); + } + // Float3 + else if (uniform.type == UniformType::Float3) + { + gl.Uniform3fv(location, (GLint)uniform.array_length, (const GLfloat*)call->material->get_value(i)); + } + // Float4 + else if (uniform.type == UniformType::Float4) + { + gl.Uniform4fv(location, (GLint)uniform.array_length, (const GLfloat*)call->material->get_value(i)); + } + // Matrix3x2 + else if (uniform.type == UniformType::Mat3x2) + { + gl.UniformMatrix3x2fv(location, (GLint)uniform.array_length, 0, (const GLfloat*)call->material->get_value(i)); + } + // Matrix4x4 + else if (uniform.type == UniformType::Mat4x4) + { + gl.UniformMatrix4fv(location, (GLint)uniform.array_length, 0, (const GLfloat*)call->material->get_value(i)); + } + } + } + + // Blend Mode + { + GLenum colorOp = gl_get_blend_func(call->blend.colorOp); + GLenum alphaOp = gl_get_blend_func(call->blend.alphaOp); + GLenum colorSrc = gl_get_blend_factor(call->blend.colorSrc); + GLenum colorDst = gl_get_blend_factor(call->blend.colorDst); + GLenum alphaSrc = gl_get_blend_factor(call->blend.alphaSrc); + GLenum alphaDst = gl_get_blend_factor(call->blend.alphaDst); + + gl.Enable(GL_BLEND); + gl.BlendEquationSeparate(colorOp, alphaOp); + gl.BlendFuncSeparate(colorSrc, colorDst, alphaSrc, alphaDst); + + gl.ColorMask( + ((int)call->blend.mask & (int)BlendMask::Red), + ((int)call->blend.mask & (int)BlendMask::Green), + ((int)call->blend.mask & (int)BlendMask::Blue), + ((int)call->blend.mask & (int)BlendMask::Alpha)); + + + unsigned char r = call->blend.rgba >> 24; + unsigned char g = call->blend.rgba >> 16; + unsigned char b = call->blend.rgba >> 8; + unsigned char a = call->blend.rgba; + + gl.BlendColor( + r / 255.0f, + g / 255.0f, + b / 255.0f, + a / 255.0f); + } + + // Depth Function + { + if (call->depth == Compare::None) + { + gl.Disable(GL_DEPTH_TEST); + } + else + { + gl.Enable(GL_DEPTH_TEST); + + switch (call->depth) + { + case Compare::None: break; + case Compare::Always: + gl.DepthFunc(GL_ALWAYS); + break; + case Compare::Equal: + gl.DepthFunc(GL_EQUAL); + break; + case Compare::Greater: + gl.DepthFunc(GL_GREATER); + break; + case Compare::GreatorOrEqual: + gl.DepthFunc(GL_GEQUAL); + break; + case Compare::Less: + gl.DepthFunc(GL_LESS); + break; + case Compare::LessOrEqual: + gl.DepthFunc(GL_LEQUAL); + break; + case Compare::Never: + gl.DepthFunc(GL_NEVER); + break; + case Compare::NotEqual: + gl.DepthFunc(GL_NOTEQUAL); + break; + } + } + } + + // Cull Mode + { + if (call->cull == Cull::None) + { + gl.Disable(GL_CULL_FACE); + } + else + { + gl.Enable(GL_CULL_FACE); + + if (call->cull == Cull::Back) + gl.CullFace(GL_BACK); + else if (call->cull == Cull::Front) + gl.CullFace(GL_FRONT); + else + gl.CullFace(GL_FRONT_AND_BACK); + } + } + + // Viewport + { + Rect viewport = call->viewport; + viewport.y = size.y - viewport.y - viewport.h; + + gl.Viewport((GLint)viewport.x, (GLint)viewport.y, (GLint)viewport.w, (GLint)viewport.h); + } + + // Scissor + { + if (!call->has_scissor) + { + gl.Disable(GL_SCISSOR_TEST); + } + else + { + Rect scissor = call->scissor; + scissor.y = size.y - scissor.y - scissor.h; + + if (scissor.w < 0) + scissor.w = 0; + if (scissor.h < 0) + scissor.h = 0; + + gl.Enable(GL_SCISSOR_TEST); + gl.Scissor((GLint)scissor.x, (GLint)scissor.y, (GLint)scissor.w, (GLint)scissor.h); + } + } + + // Draw the Mesh + { + gl.BindVertexArray(((OpenGL_Mesh*)call->mesh.get())->gl_id()); + + if (call->instance_count > 0) + { + gl.DrawElementsInstanced( + GL_TRIANGLES, + (GLint)(call->index_count), + GL_UNSIGNED_INT, + (void*)(sizeof(int) * call->index_start), + (GLint)call->instance_count); + } + else + { + gl.DrawElements( + GL_TRIANGLES, + (GLint)(call->index_count), + GL_UNSIGNED_INT, + (void*)(sizeof(int) * call->index_start)); + } + + gl.BindVertexArray(0); + } + } + + virtual void clear(const FrameBufferRef& target, uint32_t rgba) override + { + if (!target || !target->is_valid()) + { + gl.BindFramebuffer(GL_FRAMEBUFFER, 0); + } + else + { + auto framebuffer = (OpenGL_FrameBuffer*)target.get(); + if (framebuffer != nullptr) + gl.BindFramebuffer(GL_FRAMEBUFFER, framebuffer->gl_id()); + } + + unsigned char r = rgba >> 24; + unsigned char g = rgba >> 16; + unsigned char b = rgba >> 8; + unsigned char a = rgba; + + gl.Disable(GL_SCISSOR_TEST); + gl.ClearColor(r / 255.0f, g / 255.0f, b / 255.0f, a / 255.0f); + gl.Clear(GL_COLOR_BUFFER_BIT); + } + }; + + bool supported() + { + return true; + } + + Internal::GraphicsDevice* create() + { + return new OpenGL::Device(); + } + + void destroy(Internal::GraphicsDevice* device) + { + delete device; + } + } + + Internal::GraphicsDeviceInfo Internal::OpenGL_DeviceInfo = { + GfxAPI::OpenGL, + OpenGL::supported, + OpenGL::create, + OpenGL::destroy + }; +} + +#else // BLAH_USE_OPENGL + +namespace Blah +{ + namespace OpenGL + { + bool supported() { return false; } + Internal::GraphicsDevice* create() { return nullptr; } + void destroy(Internal::GraphicsDevice* device) { } + } + + Internal::GraphicsDeviceInfo Internal::OpenGL_DeviceInfo = { + GfxAPI::OpenGL, + OpenGL::supported, + OpenGL::create, + OpenGL::destroy + }; +} + +#endif // BLAH_USE_OPENGL \ No newline at end of file diff --git a/private/blah/internal/input.h b/private/blah/internal/input.h new file mode 100644 index 0000000..d5279e4 --- /dev/null +++ b/private/blah/internal/input.h @@ -0,0 +1,57 @@ +#pragma once +#include + +namespace Blah +{ + namespace Internal + { + namespace Input + { + // This is called internally by the app, and initializes the input state + void init(); + + // This is called internally by the app, and updates the input state + void frame(); + + // Call this when the Mouse moves relative to the window + void on_mouse_move(float x, float y); + + // Call this when the Mouse moves relative to the screen + void on_mouse_screen_move(float x, float y); + + // Call this when a Mouse Button is pressed + void on_mouse_down(MouseButton button); + + // Call this when a Mouse Button is released + void on_mouse_up(MouseButton button); + + // Call this when the Mouse Wheel moves + void on_mouse_wheel(Point wheel); + + // Call this when a keyboard key is pressed + void on_key_down(Key key); + + // Call this when a keyboard key is released + void on_key_up(Key key); + + // Call this on Text Input + void on_text_utf8(const char* text); + + // Call this when a Controller is connected. Note that the Name parameter must be kept valid + // until on_controller_disconnect is called with the same index. + void on_controller_connect(int index, const char* name, int isGamepad, int buttonCount, int axisCount); + + // Call this when a controller is disconnected + void on_controller_disconnect(int index); + + // Call this when a controller button is pressed + void on_button_down(int index, int button); + + // Call this when a controller button is released + void on_button_up(int index, int button); + + /// Call this when a controller axis is moved + void on_axis_move(int index, int axis, float value); + } + } +} \ No newline at end of file diff --git a/private/blah/internal/platform.h b/private/blah/internal/platform.h new file mode 100644 index 0000000..b671e19 --- /dev/null +++ b/private/blah/internal/platform.h @@ -0,0 +1,120 @@ +#pragma once +#include +#include +#include + +namespace Blah +{ + struct Config; + enum class FileMode; + + namespace Internal + { + namespace Platform + { + typedef void* FileHandle; + + // Initialize the System + bool init(const Config* config); + + // Called after the on_startup callback, but before the update loop begins + void ready(); + + // Called during shutdown + void shutdown(); + + // The time, in milliseconds, since the Application was started + uint64_t time(); + + // Called every frame + void frame(); + + // Sleeps the current thread + void sleep(int milliseconds); + + // Called to present the window contents + void present(); + + // Gets the Application Window Title in UTF-8 + const char* get_title(); + + // Sets the Application Window Title in UTF-8 + void set_title(const char* title); + + // Gets the Application Window Position, in Screen Coordinates + void get_position(int* x, int* y); + + // Sets the Application Window Position, in Screen Coordinates + void set_position(int x, int y); + + // Sets the Window Fullscreen if enabled is not 0 + void set_fullscreen(bool enabled); + + // Gets the Application Window Size, in Screen Coordinates + void get_size(int* width, int* height); + + // Sets the Application Window Size, in Screen Coordinates + void set_size(int width, int height); + + // Gets the Application Window Drawing Size, in Pixels. This may differ from the Window Size on hi-dpi displays. + void get_draw_size(int* width, int* height); + + // Gets the Desktop Content Scale. Gui should be scaled by this value + float get_content_scale(); + + // Returns the absoluate path to the directory that the application was started from + const char* app_path(); + + // Returns the absolute path to the user directory where save data and settings should be stored + const char* user_path(); + + // Returns true if a file with the given path exists + bool file_exists(const char* path); + + // Returns true if a file with the given path was deleted + bool file_delete(const char* path); + + // Returns true if a directory with the given path was successfully created + bool dir_create(const char* path); + + // Returns true if a directory with the given path exists + bool dir_exists(const char* path); + + // Returns true if a directory with the given path was deleted + bool dir_delete(const char* path); + + // enumerates a directory and appends each file to the given list + void dir_enumerate(List& list, const char* path, bool recursive); + + // opens a directory in the OS file explorer / finder + void dir_explore(const char* path); + + // Opens a file and sets the handle. returns true if the file was successfully opened + bool file_open(const char* path, FileHandle* handle, FileMode mode); + + // Returns the length of the file + int64_t file_length(FileHandle file); + + // Returns the Position of the file + int64_t file_position(FileHandle file); + + // Seeks the Position of the file and returns the new position from the start of the file + int64_t file_seek(FileHandle file, int64_t seekTo); + + // Reads a specific number of elements of a given size from the file into ptr + int64_t file_read(FileHandle file, void* ptr, int64_t size); + + // Writes a specific number of elements of the given size from ptr to the file + int64_t file_write(FileHandle file, const void* ptr, int64_t size); + + // Closes a file + void file_close(FileHandle file); + + // OpenGL Methods + void* gl_get_func(const char* name); + void* gl_context_create(); + void gl_context_make_current(void* context); + void gl_context_destroy(void* context); + } + } +} \ No newline at end of file diff --git a/private/blah/internal/platform_sdl2.cpp b/private/blah/internal/platform_sdl2.cpp new file mode 100644 index 0000000..e2d3da2 --- /dev/null +++ b/private/blah/internal/platform_sdl2.cpp @@ -0,0 +1,652 @@ +#ifdef BLAH_USE_SDL2 + +#include +#include +#include +#include +#include +#include + +#include +#include + +#if _WIN32 +// on Windows we're using the C++ API for now +#include +#include // for SetProcessDPIAware +#include // for File Reading/Writing +#include // for file explore +namespace fs = std::filesystem; +#else +// on non-Windows we use POSIX standard file system stuff +#include +#include +#include +#endif + +using namespace Blah; +using namespace Internal; + +namespace +{ + SDL_Window* window = nullptr; + SDL_Joystick* joysticks[BLAH_MAX_CONTROLLERS]; + SDL_GameController* gamepads[BLAH_MAX_CONTROLLERS]; + char* basePath = nullptr; + char* userPath = nullptr; + bool displayed = false; + + void sdl_log(void* userdata, int category, SDL_LogPriority priority, const char* message) + { + if (priority <= SDL_LOG_PRIORITY_INFO) + Log::print(message); + else if (priority <= SDL_LOG_PRIORITY_WARN) + Log::warn(message); + else + Log::error(message); + } +} + +bool Platform::init(const Config* config) +{ + // Required to call this for Windows + // I'm not sure why SDL2 doesn't do this on Windows automatically? +#if _WIN32 + SetProcessDPIAware(); +#endif + + // TODO: + // control this via some kind of config flag + SDL_LogSetAllPriority(SDL_LOG_PRIORITY_VERBOSE); + SDL_LogSetOutputFunction(sdl_log, nullptr); + + // Get SDL version + SDL_version version; + SDL_GetVersion(&version); + Log::print("SDL v%i.%i.%i", version.major, version.minor, version.patch); + + // initialize SDL + if (SDL_Init(SDL_INIT_VIDEO | SDL_INIT_TIMER | SDL_INIT_EVENTS | SDL_INIT_JOYSTICK | SDL_INIT_GAMECONTROLLER) != 0) + { + Log::error("Failed to initialize SDL2"); + return false; + } + + // GL Attributes + if (config->graphics == GfxAPI::OpenGL) + { + SDL_GL_SetAttribute(SDL_GL_CONTEXT_MAJOR_VERSION, 3); + SDL_GL_SetAttribute(SDL_GL_CONTEXT_MINOR_VERSION, 3); + SDL_GL_SetAttribute(SDL_GL_CONTEXT_PROFILE_MASK, SDL_GL_CONTEXT_PROFILE_CORE); + SDL_GL_SetAttribute(SDL_GL_CONTEXT_FLAGS, SDL_GL_CONTEXT_FORWARD_COMPATIBLE_FLAG); + SDL_GL_SetAttribute(SDL_GL_DOUBLEBUFFER, 1); + + // TODO: + // This should be controlled via the gfx api somehow? + SDL_GL_SetAttribute(SDL_GL_DEPTH_SIZE, 24); + SDL_GL_SetAttribute(SDL_GL_STENCIL_SIZE, 8); + SDL_GL_SetAttribute(SDL_GL_MULTISAMPLEBUFFERS, 1); + SDL_GL_SetAttribute(SDL_GL_MULTISAMPLESAMPLES, 4); + } + + // set up window flags + int flags = SDL_WINDOW_ALLOW_HIGHDPI | SDL_WINDOW_HIDDEN | SDL_WINDOW_RESIZABLE; + if (config->graphics == GfxAPI::OpenGL) + flags |= SDL_WINDOW_OPENGL; + + // create the window + window = SDL_CreateWindow(config->name, SDL_WINDOWPOS_CENTERED, SDL_WINDOWPOS_CENTERED, config->width, config->height, flags); + if (window == nullptr) + { + Log::error("Failed to create a Window"); + return false; + } + + // Scale Window to monitor for High DPI displays + // Other platforms do this automatically ... Windows we need to explitely do so +#if _WIN32 + { + // find the display index + int display = SDL_GetWindowDisplayIndex(window); + float ddpi, hdpi, vdpi; + if (SDL_GetDisplayDPI(display, &ddpi, &hdpi, &vdpi) == 0) + { + // scale the window up basesd on the display DPI + float hidpiRes = 96; + float dpi = (ddpi / hidpiRes); + if (dpi != 1) + { + SDL_DisplayMode mode; + SDL_GetDesktopDisplayMode(display, &mode); + SDL_SetWindowPosition(window, (int)(mode.w - config->width * dpi) / 2, (int)(mode.h - config->height * dpi) / 2); + SDL_SetWindowSize(window, (int)(config->width * dpi), (int)(config->height * dpi)); + } + } + } +#endif + + // set window properties + SDL_SetWindowResizable(window, SDL_TRUE); + SDL_SetWindowMinimumSize(window, 256, 256); + + return true; +} + +void Platform::ready() +{ + // enable V-Sync + if (App::config()->graphics == GfxAPI::OpenGL) + SDL_GL_SetSwapInterval(1); +} + +void Platform::shutdown() +{ + if (window != nullptr) + SDL_DestroyWindow(window); + window = nullptr; + displayed = false; + + if (basePath != nullptr) + SDL_free(basePath); + + if (userPath != nullptr) + SDL_free(userPath); + + SDL_Quit(); +} + +uint64_t Platform::time() +{ + return (uint64_t)SDL_GetTicks(); +} + +void Platform::frame() +{ + // update the mouse every frame + { + int winX, winY, x, y; + SDL_GetWindowPosition(window, &winX, &winY); + SDL_GetGlobalMouseState(&x, &y); + + Internal::Input::on_mouse_move((float)(x - winX), (float)(y - winY)); + Internal::Input::on_mouse_screen_move((float)x, (float)y); + } + + // poll normal events + SDL_Event event; + while (SDL_PollEvent(&event)) + { + if (event.type == SDL_QUIT) + { + auto config = App::config(); + if (config->on_exit_request != nullptr) + config->on_exit_request(); + } + // Mouse + else if (event.type == SDL_MOUSEBUTTONDOWN) + { + MouseButton btn = MouseButton::None; + if (event.button.button == SDL_BUTTON_LEFT) + btn = MouseButton::Left; + else if (event.button.button == SDL_BUTTON_RIGHT) + btn = MouseButton::Right; + else if (event.button.button == SDL_BUTTON_MIDDLE) + btn = MouseButton::Middle; + Internal::Input::on_mouse_down(btn); + } + else if (event.type == SDL_MOUSEBUTTONUP) + { + MouseButton btn = MouseButton::None; + if (event.button.button == SDL_BUTTON_LEFT) + btn = MouseButton::Left; + else if (event.button.button == SDL_BUTTON_RIGHT) + btn = MouseButton::Right; + else if (event.button.button == SDL_BUTTON_MIDDLE) + btn = MouseButton::Middle; + Internal::Input::on_mouse_up(btn); + } + else if (event.type == SDL_MOUSEWHEEL) + { + Internal::Input::on_mouse_wheel(Point(event.wheel.x, event.wheel.y)); + } + // Keyboard + else if (event.type == SDL_KEYDOWN) + { + if (event.key.repeat == 0) + Internal::Input::on_key_down((Key)event.key.keysym.scancode); + } + else if (event.type == SDL_KEYUP) + { + if (event.key.repeat == 0) + Internal::Input::on_key_up((Key)event.key.keysym.scancode); + } + else if (event.type == SDL_TEXTINPUT) + { + Internal::Input::on_text_utf8(event.text.text); + } + // Joystick Controller + else if (event.type == SDL_JOYDEVICEADDED) + { + Sint32 index = event.jdevice.which; + + if (SDL_IsGameController(index) == SDL_FALSE) + { + SDL_Joystick* ptr = joysticks[index] = SDL_JoystickOpen(index); + const char* name = SDL_JoystickName(ptr); + int button_count = SDL_JoystickNumButtons(ptr); + int axis_count = SDL_JoystickNumAxes(ptr); + + Internal::Input::on_controller_connect(index, name, 0, button_count, axis_count); + } + } + else if (event.type == SDL_JOYDEVICEREMOVED) + { + Sint32 index = event.jdevice.which; + + if (SDL_IsGameController(index) == SDL_FALSE) + { + Internal::Input::on_controller_disconnect(index); + SDL_JoystickClose(joysticks[index]); + } + } + else if (event.type == SDL_JOYBUTTONDOWN) + { + Sint32 index = event.jdevice.which; + if (SDL_IsGameController(index) == SDL_FALSE) + Internal::Input::on_button_down(index, event.jbutton.button); + } + else if (event.type == SDL_JOYBUTTONUP) + { + Sint32 index = event.jdevice.which; + if (SDL_IsGameController(index) == SDL_FALSE) + Internal::Input::on_button_up(index, event.jbutton.button); + } + else if (event.type == SDL_JOYAXISMOTION) + { + Sint32 index = event.jaxis.which; + if (SDL_IsGameController(index) == SDL_FALSE) + { + float value; + if (event.jaxis.value >= 0) + value = event.jaxis.value / 32767.0f; + else + value = event.jaxis.value / 32768.0f; + Internal::Input::on_axis_move(index, event.jaxis.axis, value); + } + } + // Gamepad Controller + else if (event.type == SDL_CONTROLLERDEVICEADDED) + { + Sint32 index = event.cdevice.which; + SDL_GameController* ptr = gamepads[index] = SDL_GameControllerOpen(index); + const char* name = SDL_GameControllerName(ptr); + Internal::Input::on_controller_connect(index, name, 1, 15, 6); + } + else if (event.type == SDL_CONTROLLERDEVICEREMOVED) + { + Sint32 index = event.cdevice.which; + Internal::Input::on_controller_disconnect(index); + SDL_GameControllerClose(gamepads[index]); + } + else if (event.type == SDL_CONTROLLERBUTTONDOWN) + { + Sint32 index = event.cbutton.which; + + int button = (int)Button::None; + if (event.cbutton.button >= 0 && event.cbutton.button < 15) + button = event.cbutton.button; // NOTE: These map directly to Engine Buttons enum! + + Internal::Input::on_button_down(index, button); + } + else if (event.type == SDL_CONTROLLERBUTTONUP) + { + Sint32 index = event.cbutton.which; + + int button = (int)Button::None; + if (event.cbutton.button >= 0 && event.cbutton.button < 15) + button = event.cbutton.button; // NOTE: These map directly to Engine Buttons enum! + + Internal::Input::on_button_up(index, button); + } + else if (event.type == SDL_CONTROLLERAXISMOTION) + { + Sint32 index = event.caxis.which; + + int axis = (int)Axis::None; + if (event.caxis.axis >= 0 && event.caxis.axis < 6) + axis = event.caxis.axis; // NOTE: These map directly to Engine Axis enum! + + float value; + if (event.caxis.value >= 0) + value = event.caxis.value / 32767.0f; + else + value = event.caxis.value / 32768.0f; + + Internal::Input::on_axis_move(index, axis, value); + } + } +} + +void Platform::sleep(int milliseconds) +{ + if (milliseconds >= 0) + SDL_Delay((uint32_t)milliseconds); +} + +void Platform::present() +{ + if (App::config()->graphics == GfxAPI::OpenGL) + { + SDL_GL_SwapWindow(window); + } + + // display the window + // this avoids a short black screen on macoS + if (!displayed) + { + SDL_ShowWindow(window); + displayed = true; + } +} + +const char* Platform::get_title() +{ + return nullptr; +} + +void Platform::set_title(const char* title) +{ + SDL_SetWindowTitle(window, title); +} + +void Platform::get_position(int* x, int* y) +{ + SDL_GetWindowPosition(window, x, y); +} + +void Platform::set_position(int x, int y) +{ + SDL_SetWindowPosition(window, x, y); +} + +void Platform::set_fullscreen(bool enabled) +{ + if (enabled) + SDL_SetWindowFullscreen(window, SDL_WINDOW_FULLSCREEN_DESKTOP); + else + SDL_SetWindowFullscreen(window, 0); +} + +void Platform::get_size(int* width, int* height) +{ + SDL_GetWindowSize(window, width, height); +} + +void Platform::set_size(int width, int height) +{ + SDL_SetWindowSize(window, width, height); +} + +void Platform::get_draw_size(int* width, int* height) +{ + auto config = App::config(); + if (config->graphics == GfxAPI::OpenGL) + { + SDL_GL_GetDrawableSize(window, width, height); + } + /* + else if (config->graphics == GfxAPI::Vulkan) + { + SDL_Vulkan_GetDrawableSize(window, width, height); + } + */ + else + { + SDL_GetWindowSize(window, width, height); + } +} + +float Platform::get_content_scale() +{ + // TODO: + // This is incorrect! but for some reason the scale + // is HUGE if I use the Display DPI on macOS :/ +#if __APPLE__ + return 2.0f; +#endif + +#if _WIN32 + float hidpiRes = 96; +#else + float hidpiRes = 72; +#endif + + int index = SDL_GetWindowDisplayIndex(window); + if (index < 0) + Log::error(SDL_GetError()); + + float ddpi, x, y; + if (SDL_GetDisplayDPI(index, &ddpi, &x, &y) != 0) + Log::error(SDL_GetError()); + + return (ddpi / hidpiRes); +} + +// FILE IO + +const char* Platform::app_path() +{ + if (basePath == nullptr) + basePath = SDL_GetBasePath(); + return basePath; +} + +const char* Platform::user_path() +{ + if (userPath == nullptr) + { + const Config* config = App::config(); + userPath = SDL_GetPrefPath(nullptr, config->name); + } + + return userPath; +} + +// Windows File System methods +#if _WIN32 + +bool Platform::file_exists(const char* path) +{ + return fs::is_regular_file(path); +} + +bool Platform::file_delete(const char* path) +{ + return fs::remove(path); +} + +bool Platform::dir_create(const char* path) +{ + std::error_code error; + return fs::create_directories(path, error); +} + +bool Platform::dir_exists(const char* path) +{ + return fs::is_directory(path); +} + +bool Platform::dir_delete(const char* path) +{ + BLAH_ERROR("not implemented"); + return false; +} + +void Platform::dir_enumerate(List& list, const char* path, bool recursive) +{ + if (fs::is_directory(path)) + { + if (recursive) + { + for (auto& p : fs::recursive_directory_iterator(path)) + { + FilePath str(p.path().string().c_str()); + list.add(str); + } + } + else + { + for (auto& p : fs::directory_iterator(path)) + { + FilePath str(p.path().string().c_str()); + list.add(str); + } + } + } +} + +void Platform::dir_explore(const char* path) +{ + ShellExecute(NULL, "open", path, NULL, NULL, SW_SHOWDEFAULT); +} + +// Non-Windows File System Methods +#else + +bool Platform::file_exists(const char* path) +{ + struct stat buffer; + return (stat(path, &buffer) == 0) && S_ISREG(buffer.st_mode); +} + +bool Platform::file_delete(const char* path) +{ + BLAH_ERROR("not implemented"); + return false; +} + +bool Platform::dir_create(const char* path) +{ + char tmp[265]; + char* p = NULL; + size_t len; + + snprintf(tmp, sizeof(tmp), "%s", path); + len = strlen(tmp); + if (tmp[len - 1] == '/') + tmp[len - 1] = 0; + for (p = tmp + 1; *p; p++) + if (*p == '/') { + *p = 0; + mkdir(tmp, S_IRWXU); + *p = '/'; + } + return mkdir(tmp, S_IRWXU) == 0; +} + +bool Platform::dir_exists(const char* path) +{ + struct stat buffer; + return (stat(path, &buffer) == 0) && S_ISDIR(buffer.st_mode); +} + +bool Platform::dir_delete(const char* path) +{ + BLAH_ERROR("not implemented"); + return false; +} + +void Platform::dir_enumerate(List& list, const char* path, bool recursive) +{ + DIR* dirp = opendir(path); + if (dirp != NULL) + { + struct dirent* dp; + while ((dp = readdir(dirp)) != NULL) + { + if (dp->d_name[0] == '.') + continue; + + FilePath subpath = FilePath(path).Append(dp->d_name); + list.Add(subpath); + + if (recursive && dp->d_type == DT_DIR) + dir_enumerate(list, subpath + "/", true); + } + closedir(dirp); + } +} + +void Platform::dir_explore(const char* path) +{ + BLAH_ERROR("not implemented"); +} + +#endif + +bool Platform::file_open(const char* path, Platform::FileHandle* handle, FileMode mode) +{ + const char* sdlMode = "rb"; + if (mode == FileMode::Write) + sdlMode = "wb"; + + auto ptr = SDL_RWFromFile(path, sdlMode); + *handle = (Platform::FileHandle)ptr; + return ptr != nullptr; +} + +int64_t Platform::file_length(Platform::FileHandle stream) +{ + return SDL_RWsize((SDL_RWops*)stream); +} + +int64_t Platform::file_position(Platform::FileHandle stream) +{ + return SDL_RWtell((SDL_RWops*)stream); +} + +int64_t Platform::file_seek(Platform::FileHandle stream, int64_t seekTo) +{ + return SDL_RWseek((SDL_RWops*)stream, seekTo, RW_SEEK_SET); +} + +int64_t Platform::file_read(Platform::FileHandle stream, void* ptr, int64_t length) +{ + return SDL_RWread((SDL_RWops*)stream, ptr, sizeof(char), length); +} + +int64_t Platform::file_write(Platform::FileHandle stream, const void* ptr, int64_t length) +{ + return SDL_RWwrite((SDL_RWops*)stream, ptr, sizeof(char), length); +} + +void Platform::file_close(Platform::FileHandle stream) +{ + if (stream != nullptr) + SDL_RWclose((SDL_RWops*)stream); +} + +void* Platform::gl_get_func(const char* name) +{ + return SDL_GL_GetProcAddress(name); +} + +void* Platform::gl_context_create() +{ + void* pointer = SDL_GL_CreateContext(window); + if (pointer == nullptr) + Log::error("SDL_GL_CreateContext failed: %s", SDL_GetError()); + return pointer; +} + +void Platform::gl_context_make_current(void* context) +{ + SDL_GL_MakeCurrent(window, context); +} + +void Platform::gl_context_destroy(void* context) +{ + SDL_GL_DeleteContext(context); +} + +#endif // BLAH_USE_SDL2 diff --git a/private/blah/third_party/stb_image.h b/private/blah/third_party/stb_image.h new file mode 100644 index 0000000..2857f05 --- /dev/null +++ b/private/blah/third_party/stb_image.h @@ -0,0 +1,7656 @@ +/* stb_image - v2.25 - public domain image loader - http://nothings.org/stb + no warranty implied; use at your own risk + + Do this: + #define STB_IMAGE_IMPLEMENTATION + before you include this file in *one* C or C++ file to create the implementation. + + // i.e. it should look like this: + #include ... + #include ... + #include ... + #define STB_IMAGE_IMPLEMENTATION + #include "stb_image.h" + + You can #define STBI_ASSERT(x) before the #include to avoid using assert.h. + And #define STBI_MALLOC, STBI_REALLOC, and STBI_FREE to avoid using malloc,realloc,free + + + QUICK NOTES: + Primarily of interest to game developers and other people who can + avoid problematic images and only need the trivial interface + + JPEG baseline & progressive (12 bpc/arithmetic not supported, same as stock IJG lib) + PNG 1/2/4/8/16-bit-per-channel + + TGA (not sure what subset, if a subset) + BMP non-1bpp, non-RLE + PSD (composited view only, no extra channels, 8/16 bit-per-channel) + + GIF (*comp always reports as 4-channel) + HDR (radiance rgbE format) + PIC (Softimage PIC) + PNM (PPM and PGM binary only) + + Animated GIF still needs a proper API, but here's one way to do it: + http://gist.github.com/urraka/685d9a6340b26b830d49 + + - decode from memory or through FILE (define STBI_NO_STDIO to remove code) + - decode from arbitrary I/O callbacks + - SIMD acceleration on x86/x64 (SSE2) and ARM (NEON) + + Full documentation under "DOCUMENTATION" below. + + +LICENSE + + See end of file for license information. + +RECENT REVISION HISTORY: + + 2.25 (2020-02-02) fix warnings + 2.24 (2020-02-02) fix warnings; thread-local failure_reason and flip_vertically + 2.23 (2019-08-11) fix clang static analysis warning + 2.22 (2019-03-04) gif fixes, fix warnings + 2.21 (2019-02-25) fix typo in comment + 2.20 (2019-02-07) support utf8 filenames in Windows; fix warnings and platform ifdefs + 2.19 (2018-02-11) fix warning + 2.18 (2018-01-30) fix warnings + 2.17 (2018-01-29) bugfix, 1-bit BMP, 16-bitness query, fix warnings + 2.16 (2017-07-23) all functions have 16-bit variants; optimizations; bugfixes + 2.15 (2017-03-18) fix png-1,2,4; all Imagenet JPGs; no runtime SSE detection on GCC + 2.14 (2017-03-03) remove deprecated STBI_JPEG_OLD; fixes for Imagenet JPGs + 2.13 (2016-12-04) experimental 16-bit API, only for PNG so far; fixes + 2.12 (2016-04-02) fix typo in 2.11 PSD fix that caused crashes + 2.11 (2016-04-02) 16-bit PNGS; enable SSE2 in non-gcc x64 + RGB-format JPEG; remove white matting in PSD; + allocate large structures on the stack; + correct channel count for PNG & BMP + 2.10 (2016-01-22) avoid warning introduced in 2.09 + 2.09 (2016-01-16) 16-bit TGA; comments in PNM files; STBI_REALLOC_SIZED + + See end of file for full revision history. + + + ============================ Contributors ========================= + + Image formats Extensions, features + Sean Barrett (jpeg, png, bmp) Jetro Lauha (stbi_info) + Nicolas Schulz (hdr, psd) Martin "SpartanJ" Golini (stbi_info) + Jonathan Dummer (tga) James "moose2000" Brown (iPhone PNG) + Jean-Marc Lienher (gif) Ben "Disch" Wenger (io callbacks) + Tom Seddon (pic) Omar Cornut (1/2/4-bit PNG) + Thatcher Ulrich (psd) Nicolas Guillemot (vertical flip) + Ken Miller (pgm, ppm) Richard Mitton (16-bit PSD) + github:urraka (animated gif) Junggon Kim (PNM comments) + Christopher Forseth (animated gif) Daniel Gibson (16-bit TGA) + socks-the-fox (16-bit PNG) + Jeremy Sawicki (handle all ImageNet JPGs) + Optimizations & bugfixes Mikhail Morozov (1-bit BMP) + Fabian "ryg" Giesen Anael Seghezzi (is-16-bit query) + Arseny Kapoulkine + John-Mark Allen + Carmelo J Fdez-Aguera + + Bug & warning fixes + Marc LeBlanc David Woo Guillaume George Martins Mozeiko + Christpher Lloyd Jerry Jansson Joseph Thomson Phil Jordan + Dave Moore Roy Eltham Hayaki Saito Nathan Reed + Won Chun Luke Graham Johan Duparc Nick Verigakis + the Horde3D community Thomas Ruf Ronny Chevalier github:rlyeh + Janez Zemva John Bartholomew Michal Cichon github:romigrou + Jonathan Blow Ken Hamada Tero Hanninen github:svdijk + Laurent Gomila Cort Stratton Sergio Gonzalez github:snagar + Aruelien Pocheville Thibault Reuille Cass Everitt github:Zelex + Ryamond Barbiero Paul Du Bois Engin Manap github:grim210 + Aldo Culquicondor Philipp Wiesemann Dale Weiler github:sammyhw + Oriol Ferrer Mesia Josh Tobin Matthew Gregan github:phprus + Julian Raschke Gregory Mullen Baldur Karlsson github:poppolopoppo + Christian Floisand Kevin Schmidt JR Smith github:darealshinji + Brad Weinberger Matvey Cherevko github:Michaelangel007 + Blazej Dariusz Roszkowski Alexander Veselov +*/ + +#ifndef STBI_INCLUDE_STB_IMAGE_H +#define STBI_INCLUDE_STB_IMAGE_H + +// DOCUMENTATION +// +// Limitations: +// - no 12-bit-per-channel JPEG +// - no JPEGs with arithmetic coding +// - GIF always returns *comp=4 +// +// Basic usage (see HDR discussion below for HDR usage): +// int x,y,n; +// unsigned char *data = stbi_load(filename, &x, &y, &n, 0); +// // ... process data if not NULL ... +// // ... x = width, y = height, n = # 8-bit components per pixel ... +// // ... replace '0' with '1'..'4' to force that many components per pixel +// // ... but 'n' will always be the number that it would have been if you said 0 +// stbi_image_free(data) +// +// Standard parameters: +// int *x -- outputs image width in pixels +// int *y -- outputs image height in pixels +// int *channels_in_file -- outputs # of image components in image file +// int desired_channels -- if non-zero, # of image components requested in result +// +// The return value from an image loader is an 'unsigned char *' which points +// to the pixel data, or NULL on an allocation failure or if the image is +// corrupt or invalid. The pixel data consists of *y scanlines of *x pixels, +// with each pixel consisting of N interleaved 8-bit components; the first +// pixel pointed to is top-left-most in the image. There is no padding between +// image scanlines or between pixels, regardless of format. The number of +// components N is 'desired_channels' if desired_channels is non-zero, or +// *channels_in_file otherwise. If desired_channels is non-zero, +// *channels_in_file has the number of components that _would_ have been +// output otherwise. E.g. if you set desired_channels to 4, you will always +// get RGBA output, but you can check *channels_in_file to see if it's trivially +// opaque because e.g. there were only 3 channels in the source image. +// +// An output image with N components has the following components interleaved +// in this order in each pixel: +// +// N=#comp components +// 1 grey +// 2 grey, alpha +// 3 red, green, blue +// 4 red, green, blue, alpha +// +// If image loading fails for any reason, the return value will be NULL, +// and *x, *y, *channels_in_file will be unchanged. The function +// stbi_failure_reason() can be queried for an extremely brief, end-user +// unfriendly explanation of why the load failed. Define STBI_NO_FAILURE_STRINGS +// to avoid compiling these strings at all, and STBI_FAILURE_USERMSG to get slightly +// more user-friendly ones. +// +// Paletted PNG, BMP, GIF, and PIC images are automatically depalettized. +// +// =========================================================================== +// +// UNICODE: +// +// If compiling for Windows and you wish to use Unicode filenames, compile +// with +// #define STBI_WINDOWS_UTF8 +// and pass utf8-encoded filenames. Call stbi_convert_wchar_to_utf8 to convert +// Windows wchar_t filenames to utf8. +// +// =========================================================================== +// +// Philosophy +// +// stb libraries are designed with the following priorities: +// +// 1. easy to use +// 2. easy to maintain +// 3. good performance +// +// Sometimes I let "good performance" creep up in priority over "easy to maintain", +// and for best performance I may provide less-easy-to-use APIs that give higher +// performance, in addition to the easy-to-use ones. Nevertheless, it's important +// to keep in mind that from the standpoint of you, a client of this library, +// all you care about is #1 and #3, and stb libraries DO NOT emphasize #3 above all. +// +// Some secondary priorities arise directly from the first two, some of which +// provide more explicit reasons why performance can't be emphasized. +// +// - Portable ("ease of use") +// - Small source code footprint ("easy to maintain") +// - No dependencies ("ease of use") +// +// =========================================================================== +// +// I/O callbacks +// +// I/O callbacks allow you to read from arbitrary sources, like packaged +// files or some other source. Data read from callbacks are processed +// through a small internal buffer (currently 128 bytes) to try to reduce +// overhead. +// +// The three functions you must define are "read" (reads some bytes of data), +// "skip" (skips some bytes of data), "eof" (reports if the stream is at the end). +// +// =========================================================================== +// +// SIMD support +// +// The JPEG decoder will try to automatically use SIMD kernels on x86 when +// supported by the compiler. For ARM Neon support, you must explicitly +// request it. +// +// (The old do-it-yourself SIMD API is no longer supported in the current +// code.) +// +// On x86, SSE2 will automatically be used when available based on a run-time +// test; if not, the generic C versions are used as a fall-back. On ARM targets, +// the typical path is to have separate builds for NEON and non-NEON devices +// (at least this is true for iOS and Android). Therefore, the NEON support is +// toggled by a build flag: define STBI_NEON to get NEON loops. +// +// If for some reason you do not want to use any of SIMD code, or if +// you have issues compiling it, you can disable it entirely by +// defining STBI_NO_SIMD. +// +// =========================================================================== +// +// HDR image support (disable by defining STBI_NO_HDR) +// +// stb_image supports loading HDR images in general, and currently the Radiance +// .HDR file format specifically. You can still load any file through the existing +// interface; if you attempt to load an HDR file, it will be automatically remapped +// to LDR, assuming gamma 2.2 and an arbitrary scale factor defaulting to 1; +// both of these constants can be reconfigured through this interface: +// +// stbi_hdr_to_ldr_gamma(2.2f); +// stbi_hdr_to_ldr_scale(1.0f); +// +// (note, do not use _inverse_ constants; stbi_image will invert them +// appropriately). +// +// Additionally, there is a new, parallel interface for loading files as +// (linear) floats to preserve the full dynamic range: +// +// float *data = stbi_loadf(filename, &x, &y, &n, 0); +// +// If you load LDR images through this interface, those images will +// be promoted to floating point values, run through the inverse of +// constants corresponding to the above: +// +// stbi_ldr_to_hdr_scale(1.0f); +// stbi_ldr_to_hdr_gamma(2.2f); +// +// Finally, given a filename (or an open file or memory block--see header +// file for details) containing image data, you can query for the "most +// appropriate" interface to use (that is, whether the image is HDR or +// not), using: +// +// stbi_is_hdr(char *filename); +// +// =========================================================================== +// +// iPhone PNG support: +// +// By default we convert iphone-formatted PNGs back to RGB, even though +// they are internally encoded differently. You can disable this conversion +// by calling stbi_convert_iphone_png_to_rgb(0), in which case +// you will always just get the native iphone "format" through (which +// is BGR stored in RGB). +// +// Call stbi_set_unpremultiply_on_load(1) as well to force a divide per +// pixel to remove any premultiplied alpha *only* if the image file explicitly +// says there's premultiplied data (currently only happens in iPhone images, +// and only if iPhone convert-to-rgb processing is on). +// +// =========================================================================== +// +// ADDITIONAL CONFIGURATION +// +// - You can suppress implementation of any of the decoders to reduce +// your code footprint by #defining one or more of the following +// symbols before creating the implementation. +// +// STBI_NO_JPEG +// STBI_NO_PNG +// STBI_NO_BMP +// STBI_NO_PSD +// STBI_NO_TGA +// STBI_NO_GIF +// STBI_NO_HDR +// STBI_NO_PIC +// STBI_NO_PNM (.ppm and .pgm) +// +// - You can request *only* certain decoders and suppress all other ones +// (this will be more forward-compatible, as addition of new decoders +// doesn't require you to disable them explicitly): +// +// STBI_ONLY_JPEG +// STBI_ONLY_PNG +// STBI_ONLY_BMP +// STBI_ONLY_PSD +// STBI_ONLY_TGA +// STBI_ONLY_GIF +// STBI_ONLY_HDR +// STBI_ONLY_PIC +// STBI_ONLY_PNM (.ppm and .pgm) +// +// - If you use STBI_NO_PNG (or _ONLY_ without PNG), and you still +// want the zlib decoder to be available, #define STBI_SUPPORT_ZLIB +// + + +#ifndef STBI_NO_STDIO +#include +#endif // STBI_NO_STDIO + +#define STBI_VERSION 1 + +enum +{ + STBI_default = 0, // only used for desired_channels + + STBI_grey = 1, + STBI_grey_alpha = 2, + STBI_rgb = 3, + STBI_rgb_alpha = 4 +}; + +#include +typedef unsigned char stbi_uc; +typedef unsigned short stbi_us; + +#ifdef __cplusplus +extern "C" { +#endif + +#ifndef STBIDEF +#ifdef STB_IMAGE_STATIC +#define STBIDEF static +#else +#define STBIDEF extern +#endif +#endif + +////////////////////////////////////////////////////////////////////////////// +// +// PRIMARY API - works on images of any type +// + +// +// load image by filename, open file, or memory buffer +// + +typedef struct +{ + int (*read) (void *user,char *data,int size); // fill 'data' with 'size' bytes. return number of bytes actually read + void (*skip) (void *user,int n); // skip the next 'n' bytes, or 'unget' the last -n bytes if negative + int (*eof) (void *user); // returns nonzero if we are at end of file/data +} stbi_io_callbacks; + +//////////////////////////////////// +// +// 8-bits-per-channel interface +// + +STBIDEF stbi_uc *stbi_load_from_memory (stbi_uc const *buffer, int len , int *x, int *y, int *channels_in_file, int desired_channels); +STBIDEF stbi_uc *stbi_load_from_callbacks(stbi_io_callbacks const *clbk , void *user, int *x, int *y, int *channels_in_file, int desired_channels); + +#ifndef STBI_NO_STDIO +STBIDEF stbi_uc *stbi_load (char const *filename, int *x, int *y, int *channels_in_file, int desired_channels); +STBIDEF stbi_uc *stbi_load_from_file (FILE *f, int *x, int *y, int *channels_in_file, int desired_channels); +// for stbi_load_from_file, file pointer is left pointing immediately after image +#endif + +#ifndef STBI_NO_GIF +STBIDEF stbi_uc *stbi_load_gif_from_memory(stbi_uc const *buffer, int len, int **delays, int *x, int *y, int *z, int *comp, int req_comp); +#endif + +#ifdef STBI_WINDOWS_UTF8 +STBIDEF int stbi_convert_wchar_to_utf8(char *buffer, size_t bufferlen, const wchar_t* input); +#endif + +//////////////////////////////////// +// +// 16-bits-per-channel interface +// + +STBIDEF stbi_us *stbi_load_16_from_memory (stbi_uc const *buffer, int len, int *x, int *y, int *channels_in_file, int desired_channels); +STBIDEF stbi_us *stbi_load_16_from_callbacks(stbi_io_callbacks const *clbk, void *user, int *x, int *y, int *channels_in_file, int desired_channels); + +#ifndef STBI_NO_STDIO +STBIDEF stbi_us *stbi_load_16 (char const *filename, int *x, int *y, int *channels_in_file, int desired_channels); +STBIDEF stbi_us *stbi_load_from_file_16(FILE *f, int *x, int *y, int *channels_in_file, int desired_channels); +#endif + +//////////////////////////////////// +// +// float-per-channel interface +// +#ifndef STBI_NO_LINEAR + STBIDEF float *stbi_loadf_from_memory (stbi_uc const *buffer, int len, int *x, int *y, int *channels_in_file, int desired_channels); + STBIDEF float *stbi_loadf_from_callbacks (stbi_io_callbacks const *clbk, void *user, int *x, int *y, int *channels_in_file, int desired_channels); + + #ifndef STBI_NO_STDIO + STBIDEF float *stbi_loadf (char const *filename, int *x, int *y, int *channels_in_file, int desired_channels); + STBIDEF float *stbi_loadf_from_file (FILE *f, int *x, int *y, int *channels_in_file, int desired_channels); + #endif +#endif + +#ifndef STBI_NO_HDR + STBIDEF void stbi_hdr_to_ldr_gamma(float gamma); + STBIDEF void stbi_hdr_to_ldr_scale(float scale); +#endif // STBI_NO_HDR + +#ifndef STBI_NO_LINEAR + STBIDEF void stbi_ldr_to_hdr_gamma(float gamma); + STBIDEF void stbi_ldr_to_hdr_scale(float scale); +#endif // STBI_NO_LINEAR + +// stbi_is_hdr is always defined, but always returns false if STBI_NO_HDR +STBIDEF int stbi_is_hdr_from_callbacks(stbi_io_callbacks const *clbk, void *user); +STBIDEF int stbi_is_hdr_from_memory(stbi_uc const *buffer, int len); +#ifndef STBI_NO_STDIO +STBIDEF int stbi_is_hdr (char const *filename); +STBIDEF int stbi_is_hdr_from_file(FILE *f); +#endif // STBI_NO_STDIO + + +// get a VERY brief reason for failure +// on most compilers (and ALL modern mainstream compilers) this is threadsafe +STBIDEF const char *stbi_failure_reason (void); + +// free the loaded image -- this is just free() +STBIDEF void stbi_image_free (void *retval_from_stbi_load); + +// get image dimensions & components without fully decoding +STBIDEF int stbi_info_from_memory(stbi_uc const *buffer, int len, int *x, int *y, int *comp); +STBIDEF int stbi_info_from_callbacks(stbi_io_callbacks const *clbk, void *user, int *x, int *y, int *comp); +STBIDEF int stbi_is_16_bit_from_memory(stbi_uc const *buffer, int len); +STBIDEF int stbi_is_16_bit_from_callbacks(stbi_io_callbacks const *clbk, void *user); + +#ifndef STBI_NO_STDIO +STBIDEF int stbi_info (char const *filename, int *x, int *y, int *comp); +STBIDEF int stbi_info_from_file (FILE *f, int *x, int *y, int *comp); +STBIDEF int stbi_is_16_bit (char const *filename); +STBIDEF int stbi_is_16_bit_from_file(FILE *f); +#endif + + + +// for image formats that explicitly notate that they have premultiplied alpha, +// we just return the colors as stored in the file. set this flag to force +// unpremultiplication. results are undefined if the unpremultiply overflow. +STBIDEF void stbi_set_unpremultiply_on_load(int flag_true_if_should_unpremultiply); + +// indicate whether we should process iphone images back to canonical format, +// or just pass them through "as-is" +STBIDEF void stbi_convert_iphone_png_to_rgb(int flag_true_if_should_convert); + +// flip the image vertically, so the first pixel in the output array is the bottom left +STBIDEF void stbi_set_flip_vertically_on_load(int flag_true_if_should_flip); + +// as above, but only applies to images loaded on the thread that calls the function +// this function is only available if your compiler supports thread-local variables; +// calling it will fail to link if your compiler doesn't +STBIDEF void stbi_set_flip_vertically_on_load_thread(int flag_true_if_should_flip); + +// ZLIB client - used by PNG, available for other purposes + +STBIDEF char *stbi_zlib_decode_malloc_guesssize(const char *buffer, int len, int initial_size, int *outlen); +STBIDEF char *stbi_zlib_decode_malloc_guesssize_headerflag(const char *buffer, int len, int initial_size, int *outlen, int parse_header); +STBIDEF char *stbi_zlib_decode_malloc(const char *buffer, int len, int *outlen); +STBIDEF int stbi_zlib_decode_buffer(char *obuffer, int olen, const char *ibuffer, int ilen); + +STBIDEF char *stbi_zlib_decode_noheader_malloc(const char *buffer, int len, int *outlen); +STBIDEF int stbi_zlib_decode_noheader_buffer(char *obuffer, int olen, const char *ibuffer, int ilen); + + +#ifdef __cplusplus +} +#endif + +// +// +//// end header file ///////////////////////////////////////////////////// +#endif // STBI_INCLUDE_STB_IMAGE_H + +#ifdef STB_IMAGE_IMPLEMENTATION + +#if defined(STBI_ONLY_JPEG) || defined(STBI_ONLY_PNG) || defined(STBI_ONLY_BMP) \ + || defined(STBI_ONLY_TGA) || defined(STBI_ONLY_GIF) || defined(STBI_ONLY_PSD) \ + || defined(STBI_ONLY_HDR) || defined(STBI_ONLY_PIC) || defined(STBI_ONLY_PNM) \ + || defined(STBI_ONLY_ZLIB) + #ifndef STBI_ONLY_JPEG + #define STBI_NO_JPEG + #endif + #ifndef STBI_ONLY_PNG + #define STBI_NO_PNG + #endif + #ifndef STBI_ONLY_BMP + #define STBI_NO_BMP + #endif + #ifndef STBI_ONLY_PSD + #define STBI_NO_PSD + #endif + #ifndef STBI_ONLY_TGA + #define STBI_NO_TGA + #endif + #ifndef STBI_ONLY_GIF + #define STBI_NO_GIF + #endif + #ifndef STBI_ONLY_HDR + #define STBI_NO_HDR + #endif + #ifndef STBI_ONLY_PIC + #define STBI_NO_PIC + #endif + #ifndef STBI_ONLY_PNM + #define STBI_NO_PNM + #endif +#endif + +#if defined(STBI_NO_PNG) && !defined(STBI_SUPPORT_ZLIB) && !defined(STBI_NO_ZLIB) +#define STBI_NO_ZLIB +#endif + + +#include +#include // ptrdiff_t on osx +#include +#include +#include + +#if !defined(STBI_NO_LINEAR) || !defined(STBI_NO_HDR) +#include // ldexp, pow +#endif + +#ifndef STBI_NO_STDIO +#include +#endif + +#ifndef STBI_ASSERT +#include +#define STBI_ASSERT(x) assert(x) +#endif + +#ifdef __cplusplus +#define STBI_EXTERN extern "C" +#else +#define STBI_EXTERN extern +#endif + + +#ifndef _MSC_VER + #ifdef __cplusplus + #define stbi_inline inline + #else + #define stbi_inline + #endif +#else + #define stbi_inline __forceinline +#endif + +#ifndef STBI_NO_THREAD_LOCALS + #if defined(__cplusplus) && __cplusplus >= 201103L + #define STBI_THREAD_LOCAL thread_local + #elif defined (__STDC_VERSION__) && __STDC_VERSION__ >= 201112L + #define STBI_THREAD_LOCAL _Thread_local + #elif defined(__GNUC__) + #define STBI_THREAD_LOCAL __thread + #elif defined(_MSC_VER) + #define STBI_THREAD_LOCAL __declspec(thread) +#endif +#endif + +#ifdef _MSC_VER +typedef unsigned short stbi__uint16; +typedef signed short stbi__int16; +typedef unsigned int stbi__uint32; +typedef signed int stbi__int32; +#else +#include +typedef uint16_t stbi__uint16; +typedef int16_t stbi__int16; +typedef uint32_t stbi__uint32; +typedef int32_t stbi__int32; +#endif + +// should produce compiler error if size is wrong +typedef unsigned char validate_uint32[sizeof(stbi__uint32)==4 ? 1 : -1]; + +#ifdef _MSC_VER +#define STBI_NOTUSED(v) (void)(v) +#else +#define STBI_NOTUSED(v) (void)sizeof(v) +#endif + +#ifdef _MSC_VER +#define STBI_HAS_LROTL +#endif + +#ifdef STBI_HAS_LROTL + #define stbi_lrot(x,y) _lrotl(x,y) +#else + #define stbi_lrot(x,y) (((x) << (y)) | ((x) >> (32 - (y)))) +#endif + +#if defined(STBI_MALLOC) && defined(STBI_FREE) && (defined(STBI_REALLOC) || defined(STBI_REALLOC_SIZED)) +// ok +#elif !defined(STBI_MALLOC) && !defined(STBI_FREE) && !defined(STBI_REALLOC) && !defined(STBI_REALLOC_SIZED) +// ok +#else +#error "Must define all or none of STBI_MALLOC, STBI_FREE, and STBI_REALLOC (or STBI_REALLOC_SIZED)." +#endif + +#ifndef STBI_MALLOC +#define STBI_MALLOC(sz) malloc(sz) +#define STBI_REALLOC(p,newsz) realloc(p,newsz) +#define STBI_FREE(p) free(p) +#endif + +#ifndef STBI_REALLOC_SIZED +#define STBI_REALLOC_SIZED(p,oldsz,newsz) STBI_REALLOC(p,newsz) +#endif + +// x86/x64 detection +#if defined(__x86_64__) || defined(_M_X64) +#define STBI__X64_TARGET +#elif defined(__i386) || defined(_M_IX86) +#define STBI__X86_TARGET +#endif + +#if defined(__GNUC__) && defined(STBI__X86_TARGET) && !defined(__SSE2__) && !defined(STBI_NO_SIMD) +// gcc doesn't support sse2 intrinsics unless you compile with -msse2, +// which in turn means it gets to use SSE2 everywhere. This is unfortunate, +// but previous attempts to provide the SSE2 functions with runtime +// detection caused numerous issues. The way architecture extensions are +// exposed in GCC/Clang is, sadly, not really suited for one-file libs. +// New behavior: if compiled with -msse2, we use SSE2 without any +// detection; if not, we don't use it at all. +#define STBI_NO_SIMD +#endif + +#if defined(__MINGW32__) && defined(STBI__X86_TARGET) && !defined(STBI_MINGW_ENABLE_SSE2) && !defined(STBI_NO_SIMD) +// Note that __MINGW32__ doesn't actually mean 32-bit, so we have to avoid STBI__X64_TARGET +// +// 32-bit MinGW wants ESP to be 16-byte aligned, but this is not in the +// Windows ABI and VC++ as well as Windows DLLs don't maintain that invariant. +// As a result, enabling SSE2 on 32-bit MinGW is dangerous when not +// simultaneously enabling "-mstackrealign". +// +// See https://github.com/nothings/stb/issues/81 for more information. +// +// So default to no SSE2 on 32-bit MinGW. If you've read this far and added +// -mstackrealign to your build settings, feel free to #define STBI_MINGW_ENABLE_SSE2. +#define STBI_NO_SIMD +#endif + +#if !defined(STBI_NO_SIMD) && (defined(STBI__X86_TARGET) || defined(STBI__X64_TARGET)) +#define STBI_SSE2 +#include + +#ifdef _MSC_VER + +#if _MSC_VER >= 1400 // not VC6 +#include // __cpuid +static int stbi__cpuid3(void) +{ + int info[4]; + __cpuid(info,1); + return info[3]; +} +#else +static int stbi__cpuid3(void) +{ + int res; + __asm { + mov eax,1 + cpuid + mov res,edx + } + return res; +} +#endif + +#define STBI_SIMD_ALIGN(type, name) __declspec(align(16)) type name + +#if !defined(STBI_NO_JPEG) && defined(STBI_SSE2) +static int stbi__sse2_available(void) +{ + int info3 = stbi__cpuid3(); + return ((info3 >> 26) & 1) != 0; +} +#endif + +#else // assume GCC-style if not VC++ +#define STBI_SIMD_ALIGN(type, name) type name __attribute__((aligned(16))) + +#if !defined(STBI_NO_JPEG) && defined(STBI_SSE2) +static int stbi__sse2_available(void) +{ + // If we're even attempting to compile this on GCC/Clang, that means + // -msse2 is on, which means the compiler is allowed to use SSE2 + // instructions at will, and so are we. + return 1; +} +#endif + +#endif +#endif + +// ARM NEON +#if defined(STBI_NO_SIMD) && defined(STBI_NEON) +#undef STBI_NEON +#endif + +#ifdef STBI_NEON +#include +// assume GCC or Clang on ARM targets +#define STBI_SIMD_ALIGN(type, name) type name __attribute__((aligned(16))) +#endif + +#ifndef STBI_SIMD_ALIGN +#define STBI_SIMD_ALIGN(type, name) type name +#endif + +/////////////////////////////////////////////// +// +// stbi__context struct and start_xxx functions + +// stbi__context structure is our basic context used by all images, so it +// contains all the IO context, plus some basic image information +typedef struct +{ + stbi__uint32 img_x, img_y; + int img_n, img_out_n; + + stbi_io_callbacks io; + void *io_user_data; + + int read_from_callbacks; + int buflen; + stbi_uc buffer_start[128]; + + stbi_uc *img_buffer, *img_buffer_end; + stbi_uc *img_buffer_original, *img_buffer_original_end; +} stbi__context; + + +static void stbi__refill_buffer(stbi__context *s); + +// initialize a memory-decode context +static void stbi__start_mem(stbi__context *s, stbi_uc const *buffer, int len) +{ + s->io.read = NULL; + s->read_from_callbacks = 0; + s->img_buffer = s->img_buffer_original = (stbi_uc *) buffer; + s->img_buffer_end = s->img_buffer_original_end = (stbi_uc *) buffer+len; +} + +// initialize a callback-based context +static void stbi__start_callbacks(stbi__context *s, stbi_io_callbacks *c, void *user) +{ + s->io = *c; + s->io_user_data = user; + s->buflen = sizeof(s->buffer_start); + s->read_from_callbacks = 1; + s->img_buffer_original = s->buffer_start; + stbi__refill_buffer(s); + s->img_buffer_original_end = s->img_buffer_end; +} + +#ifndef STBI_NO_STDIO + +static int stbi__stdio_read(void *user, char *data, int size) +{ + return (int) fread(data,1,size,(FILE*) user); +} + +static void stbi__stdio_skip(void *user, int n) +{ + fseek((FILE*) user, n, SEEK_CUR); +} + +static int stbi__stdio_eof(void *user) +{ + return feof((FILE*) user); +} + +static stbi_io_callbacks stbi__stdio_callbacks = +{ + stbi__stdio_read, + stbi__stdio_skip, + stbi__stdio_eof, +}; + +static void stbi__start_file(stbi__context *s, FILE *f) +{ + stbi__start_callbacks(s, &stbi__stdio_callbacks, (void *) f); +} + +//static void stop_file(stbi__context *s) { } + +#endif // !STBI_NO_STDIO + +static void stbi__rewind(stbi__context *s) +{ + // conceptually rewind SHOULD rewind to the beginning of the stream, + // but we just rewind to the beginning of the initial buffer, because + // we only use it after doing 'test', which only ever looks at at most 92 bytes + s->img_buffer = s->img_buffer_original; + s->img_buffer_end = s->img_buffer_original_end; +} + +enum +{ + STBI_ORDER_RGB, + STBI_ORDER_BGR +}; + +typedef struct +{ + int bits_per_channel; + int num_channels; + int channel_order; +} stbi__result_info; + +#ifndef STBI_NO_JPEG +static int stbi__jpeg_test(stbi__context *s); +static void *stbi__jpeg_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri); +static int stbi__jpeg_info(stbi__context *s, int *x, int *y, int *comp); +#endif + +#ifndef STBI_NO_PNG +static int stbi__png_test(stbi__context *s); +static void *stbi__png_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri); +static int stbi__png_info(stbi__context *s, int *x, int *y, int *comp); +static int stbi__png_is16(stbi__context *s); +#endif + +#ifndef STBI_NO_BMP +static int stbi__bmp_test(stbi__context *s); +static void *stbi__bmp_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri); +static int stbi__bmp_info(stbi__context *s, int *x, int *y, int *comp); +#endif + +#ifndef STBI_NO_TGA +static int stbi__tga_test(stbi__context *s); +static void *stbi__tga_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri); +static int stbi__tga_info(stbi__context *s, int *x, int *y, int *comp); +#endif + +#ifndef STBI_NO_PSD +static int stbi__psd_test(stbi__context *s); +static void *stbi__psd_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri, int bpc); +static int stbi__psd_info(stbi__context *s, int *x, int *y, int *comp); +static int stbi__psd_is16(stbi__context *s); +#endif + +#ifndef STBI_NO_HDR +static int stbi__hdr_test(stbi__context *s); +static float *stbi__hdr_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri); +static int stbi__hdr_info(stbi__context *s, int *x, int *y, int *comp); +#endif + +#ifndef STBI_NO_PIC +static int stbi__pic_test(stbi__context *s); +static void *stbi__pic_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri); +static int stbi__pic_info(stbi__context *s, int *x, int *y, int *comp); +#endif + +#ifndef STBI_NO_GIF +static int stbi__gif_test(stbi__context *s); +static void *stbi__gif_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri); +static void *stbi__load_gif_main(stbi__context *s, int **delays, int *x, int *y, int *z, int *comp, int req_comp); +static int stbi__gif_info(stbi__context *s, int *x, int *y, int *comp); +#endif + +#ifndef STBI_NO_PNM +static int stbi__pnm_test(stbi__context *s); +static void *stbi__pnm_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri); +static int stbi__pnm_info(stbi__context *s, int *x, int *y, int *comp); +#endif + +static +#ifdef STBI_THREAD_LOCAL +STBI_THREAD_LOCAL +#endif +const char *stbi__g_failure_reason; + +STBIDEF const char *stbi_failure_reason(void) +{ + return stbi__g_failure_reason; +} + +#ifndef STBI_NO_FAILURE_STRINGS +static int stbi__err(const char *str) +{ + stbi__g_failure_reason = str; + return 0; +} +#endif + +static void *stbi__malloc(size_t size) +{ + return STBI_MALLOC(size); +} + +// stb_image uses ints pervasively, including for offset calculations. +// therefore the largest decoded image size we can support with the +// current code, even on 64-bit targets, is INT_MAX. this is not a +// significant limitation for the intended use case. +// +// we do, however, need to make sure our size calculations don't +// overflow. hence a few helper functions for size calculations that +// multiply integers together, making sure that they're non-negative +// and no overflow occurs. + +// return 1 if the sum is valid, 0 on overflow. +// negative terms are considered invalid. +static int stbi__addsizes_valid(int a, int b) +{ + if (b < 0) return 0; + // now 0 <= b <= INT_MAX, hence also + // 0 <= INT_MAX - b <= INTMAX. + // And "a + b <= INT_MAX" (which might overflow) is the + // same as a <= INT_MAX - b (no overflow) + return a <= INT_MAX - b; +} + +// returns 1 if the product is valid, 0 on overflow. +// negative factors are considered invalid. +static int stbi__mul2sizes_valid(int a, int b) +{ + if (a < 0 || b < 0) return 0; + if (b == 0) return 1; // mul-by-0 is always safe + // portable way to check for no overflows in a*b + return a <= INT_MAX/b; +} + +#if !defined(STBI_NO_JPEG) || !defined(STBI_NO_PNG) || !defined(STBI_NO_TGA) || !defined(STBI_NO_HDR) +// returns 1 if "a*b + add" has no negative terms/factors and doesn't overflow +static int stbi__mad2sizes_valid(int a, int b, int add) +{ + return stbi__mul2sizes_valid(a, b) && stbi__addsizes_valid(a*b, add); +} +#endif + +// returns 1 if "a*b*c + add" has no negative terms/factors and doesn't overflow +static int stbi__mad3sizes_valid(int a, int b, int c, int add) +{ + return stbi__mul2sizes_valid(a, b) && stbi__mul2sizes_valid(a*b, c) && + stbi__addsizes_valid(a*b*c, add); +} + +// returns 1 if "a*b*c*d + add" has no negative terms/factors and doesn't overflow +#if !defined(STBI_NO_LINEAR) || !defined(STBI_NO_HDR) +static int stbi__mad4sizes_valid(int a, int b, int c, int d, int add) +{ + return stbi__mul2sizes_valid(a, b) && stbi__mul2sizes_valid(a*b, c) && + stbi__mul2sizes_valid(a*b*c, d) && stbi__addsizes_valid(a*b*c*d, add); +} +#endif + +#if !defined(STBI_NO_JPEG) || !defined(STBI_NO_PNG) || !defined(STBI_NO_TGA) || !defined(STBI_NO_HDR) +// mallocs with size overflow checking +static void *stbi__malloc_mad2(int a, int b, int add) +{ + if (!stbi__mad2sizes_valid(a, b, add)) return NULL; + return stbi__malloc(a*b + add); +} +#endif + +static void *stbi__malloc_mad3(int a, int b, int c, int add) +{ + if (!stbi__mad3sizes_valid(a, b, c, add)) return NULL; + return stbi__malloc(a*b*c + add); +} + +#if !defined(STBI_NO_LINEAR) || !defined(STBI_NO_HDR) +static void *stbi__malloc_mad4(int a, int b, int c, int d, int add) +{ + if (!stbi__mad4sizes_valid(a, b, c, d, add)) return NULL; + return stbi__malloc(a*b*c*d + add); +} +#endif + +// stbi__err - error +// stbi__errpf - error returning pointer to float +// stbi__errpuc - error returning pointer to unsigned char + +#ifdef STBI_NO_FAILURE_STRINGS + #define stbi__err(x,y) 0 +#elif defined(STBI_FAILURE_USERMSG) + #define stbi__err(x,y) stbi__err(y) +#else + #define stbi__err(x,y) stbi__err(x) +#endif + +#define stbi__errpf(x,y) ((float *)(size_t) (stbi__err(x,y)?NULL:NULL)) +#define stbi__errpuc(x,y) ((unsigned char *)(size_t) (stbi__err(x,y)?NULL:NULL)) + +STBIDEF void stbi_image_free(void *retval_from_stbi_load) +{ + STBI_FREE(retval_from_stbi_load); +} + +#ifndef STBI_NO_LINEAR +static float *stbi__ldr_to_hdr(stbi_uc *data, int x, int y, int comp); +#endif + +#ifndef STBI_NO_HDR +static stbi_uc *stbi__hdr_to_ldr(float *data, int x, int y, int comp); +#endif + +static int stbi__vertically_flip_on_load_global = 0; + +STBIDEF void stbi_set_flip_vertically_on_load(int flag_true_if_should_flip) +{ + stbi__vertically_flip_on_load_global = flag_true_if_should_flip; +} + +#ifndef STBI_THREAD_LOCAL +#define stbi__vertically_flip_on_load stbi__vertically_flip_on_load_global +#else +static STBI_THREAD_LOCAL int stbi__vertically_flip_on_load_local, stbi__vertically_flip_on_load_set; + +STBIDEF void stbi_set_flip_vertically_on_load_thread(int flag_true_if_should_flip) +{ + stbi__vertically_flip_on_load_local = flag_true_if_should_flip; + stbi__vertically_flip_on_load_set = 1; +} + +#define stbi__vertically_flip_on_load (stbi__vertically_flip_on_load_set \ + ? stbi__vertically_flip_on_load_local \ + : stbi__vertically_flip_on_load_global) +#endif // STBI_THREAD_LOCAL + +static void *stbi__load_main(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri, int bpc) +{ + memset(ri, 0, sizeof(*ri)); // make sure it's initialized if we add new fields + ri->bits_per_channel = 8; // default is 8 so most paths don't have to be changed + ri->channel_order = STBI_ORDER_RGB; // all current input & output are this, but this is here so we can add BGR order + ri->num_channels = 0; + + #ifndef STBI_NO_JPEG + if (stbi__jpeg_test(s)) return stbi__jpeg_load(s,x,y,comp,req_comp, ri); + #endif + #ifndef STBI_NO_PNG + if (stbi__png_test(s)) return stbi__png_load(s,x,y,comp,req_comp, ri); + #endif + #ifndef STBI_NO_BMP + if (stbi__bmp_test(s)) return stbi__bmp_load(s,x,y,comp,req_comp, ri); + #endif + #ifndef STBI_NO_GIF + if (stbi__gif_test(s)) return stbi__gif_load(s,x,y,comp,req_comp, ri); + #endif + #ifndef STBI_NO_PSD + if (stbi__psd_test(s)) return stbi__psd_load(s,x,y,comp,req_comp, ri, bpc); + #else + STBI_NOTUSED(bpc); + #endif + #ifndef STBI_NO_PIC + if (stbi__pic_test(s)) return stbi__pic_load(s,x,y,comp,req_comp, ri); + #endif + #ifndef STBI_NO_PNM + if (stbi__pnm_test(s)) return stbi__pnm_load(s,x,y,comp,req_comp, ri); + #endif + + #ifndef STBI_NO_HDR + if (stbi__hdr_test(s)) { + float *hdr = stbi__hdr_load(s, x,y,comp,req_comp, ri); + return stbi__hdr_to_ldr(hdr, *x, *y, req_comp ? req_comp : *comp); + } + #endif + + #ifndef STBI_NO_TGA + // test tga last because it's a crappy test! + if (stbi__tga_test(s)) + return stbi__tga_load(s,x,y,comp,req_comp, ri); + #endif + + return stbi__errpuc("unknown image type", "Image not of any known type, or corrupt"); +} + +static stbi_uc *stbi__convert_16_to_8(stbi__uint16 *orig, int w, int h, int channels) +{ + int i; + int img_len = w * h * channels; + stbi_uc *reduced; + + reduced = (stbi_uc *) stbi__malloc(img_len); + if (reduced == NULL) return stbi__errpuc("outofmem", "Out of memory"); + + for (i = 0; i < img_len; ++i) + reduced[i] = (stbi_uc)((orig[i] >> 8) & 0xFF); // top half of each byte is sufficient approx of 16->8 bit scaling + + STBI_FREE(orig); + return reduced; +} + +static stbi__uint16 *stbi__convert_8_to_16(stbi_uc *orig, int w, int h, int channels) +{ + int i; + int img_len = w * h * channels; + stbi__uint16 *enlarged; + + enlarged = (stbi__uint16 *) stbi__malloc(img_len*2); + if (enlarged == NULL) return (stbi__uint16 *) stbi__errpuc("outofmem", "Out of memory"); + + for (i = 0; i < img_len; ++i) + enlarged[i] = (stbi__uint16)((orig[i] << 8) + orig[i]); // replicate to high and low byte, maps 0->0, 255->0xffff + + STBI_FREE(orig); + return enlarged; +} + +static void stbi__vertical_flip(void *image, int w, int h, int bytes_per_pixel) +{ + int row; + size_t bytes_per_row = (size_t)w * bytes_per_pixel; + stbi_uc temp[2048]; + stbi_uc *bytes = (stbi_uc *)image; + + for (row = 0; row < (h>>1); row++) { + stbi_uc *row0 = bytes + row*bytes_per_row; + stbi_uc *row1 = bytes + (h - row - 1)*bytes_per_row; + // swap row0 with row1 + size_t bytes_left = bytes_per_row; + while (bytes_left) { + size_t bytes_copy = (bytes_left < sizeof(temp)) ? bytes_left : sizeof(temp); + memcpy(temp, row0, bytes_copy); + memcpy(row0, row1, bytes_copy); + memcpy(row1, temp, bytes_copy); + row0 += bytes_copy; + row1 += bytes_copy; + bytes_left -= bytes_copy; + } + } +} + +#ifndef STBI_NO_GIF +static void stbi__vertical_flip_slices(void *image, int w, int h, int z, int bytes_per_pixel) +{ + int slice; + int slice_size = w * h * bytes_per_pixel; + + stbi_uc *bytes = (stbi_uc *)image; + for (slice = 0; slice < z; ++slice) { + stbi__vertical_flip(bytes, w, h, bytes_per_pixel); + bytes += slice_size; + } +} +#endif + +static unsigned char *stbi__load_and_postprocess_8bit(stbi__context *s, int *x, int *y, int *comp, int req_comp) +{ + stbi__result_info ri; + void *result = stbi__load_main(s, x, y, comp, req_comp, &ri, 8); + + if (result == NULL) + return NULL; + + if (ri.bits_per_channel != 8) { + STBI_ASSERT(ri.bits_per_channel == 16); + result = stbi__convert_16_to_8((stbi__uint16 *) result, *x, *y, req_comp == 0 ? *comp : req_comp); + ri.bits_per_channel = 8; + } + + // @TODO: move stbi__convert_format to here + + if (stbi__vertically_flip_on_load) { + int channels = req_comp ? req_comp : *comp; + stbi__vertical_flip(result, *x, *y, channels * sizeof(stbi_uc)); + } + + return (unsigned char *) result; +} + +static stbi__uint16 *stbi__load_and_postprocess_16bit(stbi__context *s, int *x, int *y, int *comp, int req_comp) +{ + stbi__result_info ri; + void *result = stbi__load_main(s, x, y, comp, req_comp, &ri, 16); + + if (result == NULL) + return NULL; + + if (ri.bits_per_channel != 16) { + STBI_ASSERT(ri.bits_per_channel == 8); + result = stbi__convert_8_to_16((stbi_uc *) result, *x, *y, req_comp == 0 ? *comp : req_comp); + ri.bits_per_channel = 16; + } + + // @TODO: move stbi__convert_format16 to here + // @TODO: special case RGB-to-Y (and RGBA-to-YA) for 8-bit-to-16-bit case to keep more precision + + if (stbi__vertically_flip_on_load) { + int channels = req_comp ? req_comp : *comp; + stbi__vertical_flip(result, *x, *y, channels * sizeof(stbi__uint16)); + } + + return (stbi__uint16 *) result; +} + +#if !defined(STBI_NO_HDR) && !defined(STBI_NO_LINEAR) +static void stbi__float_postprocess(float *result, int *x, int *y, int *comp, int req_comp) +{ + if (stbi__vertically_flip_on_load && result != NULL) { + int channels = req_comp ? req_comp : *comp; + stbi__vertical_flip(result, *x, *y, channels * sizeof(float)); + } +} +#endif + +#ifndef STBI_NO_STDIO + +#if defined(_MSC_VER) && defined(STBI_WINDOWS_UTF8) +STBI_EXTERN __declspec(dllimport) int __stdcall MultiByteToWideChar(unsigned int cp, unsigned long flags, const char *str, int cbmb, wchar_t *widestr, int cchwide); +STBI_EXTERN __declspec(dllimport) int __stdcall WideCharToMultiByte(unsigned int cp, unsigned long flags, const wchar_t *widestr, int cchwide, char *str, int cbmb, const char *defchar, int *used_default); +#endif + +#if defined(_MSC_VER) && defined(STBI_WINDOWS_UTF8) +STBIDEF int stbi_convert_wchar_to_utf8(char *buffer, size_t bufferlen, const wchar_t* input) +{ + return WideCharToMultiByte(65001 /* UTF8 */, 0, input, -1, buffer, (int) bufferlen, NULL, NULL); +} +#endif + +static FILE *stbi__fopen(char const *filename, char const *mode) +{ + FILE *f; +#if defined(_MSC_VER) && defined(STBI_WINDOWS_UTF8) + wchar_t wMode[64]; + wchar_t wFilename[1024]; + if (0 == MultiByteToWideChar(65001 /* UTF8 */, 0, filename, -1, wFilename, sizeof(wFilename))) + return 0; + + if (0 == MultiByteToWideChar(65001 /* UTF8 */, 0, mode, -1, wMode, sizeof(wMode))) + return 0; + +#if _MSC_VER >= 1400 + if (0 != _wfopen_s(&f, wFilename, wMode)) + f = 0; +#else + f = _wfopen(wFilename, wMode); +#endif + +#elif defined(_MSC_VER) && _MSC_VER >= 1400 + if (0 != fopen_s(&f, filename, mode)) + f=0; +#else + f = fopen(filename, mode); +#endif + return f; +} + + +STBIDEF stbi_uc *stbi_load(char const *filename, int *x, int *y, int *comp, int req_comp) +{ + FILE *f = stbi__fopen(filename, "rb"); + unsigned char *result; + if (!f) return stbi__errpuc("can't fopen", "Unable to open file"); + result = stbi_load_from_file(f,x,y,comp,req_comp); + fclose(f); + return result; +} + +STBIDEF stbi_uc *stbi_load_from_file(FILE *f, int *x, int *y, int *comp, int req_comp) +{ + unsigned char *result; + stbi__context s; + stbi__start_file(&s,f); + result = stbi__load_and_postprocess_8bit(&s,x,y,comp,req_comp); + if (result) { + // need to 'unget' all the characters in the IO buffer + fseek(f, - (int) (s.img_buffer_end - s.img_buffer), SEEK_CUR); + } + return result; +} + +STBIDEF stbi__uint16 *stbi_load_from_file_16(FILE *f, int *x, int *y, int *comp, int req_comp) +{ + stbi__uint16 *result; + stbi__context s; + stbi__start_file(&s,f); + result = stbi__load_and_postprocess_16bit(&s,x,y,comp,req_comp); + if (result) { + // need to 'unget' all the characters in the IO buffer + fseek(f, - (int) (s.img_buffer_end - s.img_buffer), SEEK_CUR); + } + return result; +} + +STBIDEF stbi_us *stbi_load_16(char const *filename, int *x, int *y, int *comp, int req_comp) +{ + FILE *f = stbi__fopen(filename, "rb"); + stbi__uint16 *result; + if (!f) return (stbi_us *) stbi__errpuc("can't fopen", "Unable to open file"); + result = stbi_load_from_file_16(f,x,y,comp,req_comp); + fclose(f); + return result; +} + + +#endif //!STBI_NO_STDIO + +STBIDEF stbi_us *stbi_load_16_from_memory(stbi_uc const *buffer, int len, int *x, int *y, int *channels_in_file, int desired_channels) +{ + stbi__context s; + stbi__start_mem(&s,buffer,len); + return stbi__load_and_postprocess_16bit(&s,x,y,channels_in_file,desired_channels); +} + +STBIDEF stbi_us *stbi_load_16_from_callbacks(stbi_io_callbacks const *clbk, void *user, int *x, int *y, int *channels_in_file, int desired_channels) +{ + stbi__context s; + stbi__start_callbacks(&s, (stbi_io_callbacks *)clbk, user); + return stbi__load_and_postprocess_16bit(&s,x,y,channels_in_file,desired_channels); +} + +STBIDEF stbi_uc *stbi_load_from_memory(stbi_uc const *buffer, int len, int *x, int *y, int *comp, int req_comp) +{ + stbi__context s; + stbi__start_mem(&s,buffer,len); + return stbi__load_and_postprocess_8bit(&s,x,y,comp,req_comp); +} + +STBIDEF stbi_uc *stbi_load_from_callbacks(stbi_io_callbacks const *clbk, void *user, int *x, int *y, int *comp, int req_comp) +{ + stbi__context s; + stbi__start_callbacks(&s, (stbi_io_callbacks *) clbk, user); + return stbi__load_and_postprocess_8bit(&s,x,y,comp,req_comp); +} + +#ifndef STBI_NO_GIF +STBIDEF stbi_uc *stbi_load_gif_from_memory(stbi_uc const *buffer, int len, int **delays, int *x, int *y, int *z, int *comp, int req_comp) +{ + unsigned char *result; + stbi__context s; + stbi__start_mem(&s,buffer,len); + + result = (unsigned char*) stbi__load_gif_main(&s, delays, x, y, z, comp, req_comp); + if (stbi__vertically_flip_on_load) { + stbi__vertical_flip_slices( result, *x, *y, *z, *comp ); + } + + return result; +} +#endif + +#ifndef STBI_NO_LINEAR +static float *stbi__loadf_main(stbi__context *s, int *x, int *y, int *comp, int req_comp) +{ + unsigned char *data; + #ifndef STBI_NO_HDR + if (stbi__hdr_test(s)) { + stbi__result_info ri; + float *hdr_data = stbi__hdr_load(s,x,y,comp,req_comp, &ri); + if (hdr_data) + stbi__float_postprocess(hdr_data,x,y,comp,req_comp); + return hdr_data; + } + #endif + data = stbi__load_and_postprocess_8bit(s, x, y, comp, req_comp); + if (data) + return stbi__ldr_to_hdr(data, *x, *y, req_comp ? req_comp : *comp); + return stbi__errpf("unknown image type", "Image not of any known type, or corrupt"); +} + +STBIDEF float *stbi_loadf_from_memory(stbi_uc const *buffer, int len, int *x, int *y, int *comp, int req_comp) +{ + stbi__context s; + stbi__start_mem(&s,buffer,len); + return stbi__loadf_main(&s,x,y,comp,req_comp); +} + +STBIDEF float *stbi_loadf_from_callbacks(stbi_io_callbacks const *clbk, void *user, int *x, int *y, int *comp, int req_comp) +{ + stbi__context s; + stbi__start_callbacks(&s, (stbi_io_callbacks *) clbk, user); + return stbi__loadf_main(&s,x,y,comp,req_comp); +} + +#ifndef STBI_NO_STDIO +STBIDEF float *stbi_loadf(char const *filename, int *x, int *y, int *comp, int req_comp) +{ + float *result; + FILE *f = stbi__fopen(filename, "rb"); + if (!f) return stbi__errpf("can't fopen", "Unable to open file"); + result = stbi_loadf_from_file(f,x,y,comp,req_comp); + fclose(f); + return result; +} + +STBIDEF float *stbi_loadf_from_file(FILE *f, int *x, int *y, int *comp, int req_comp) +{ + stbi__context s; + stbi__start_file(&s,f); + return stbi__loadf_main(&s,x,y,comp,req_comp); +} +#endif // !STBI_NO_STDIO + +#endif // !STBI_NO_LINEAR + +// these is-hdr-or-not is defined independent of whether STBI_NO_LINEAR is +// defined, for API simplicity; if STBI_NO_LINEAR is defined, it always +// reports false! + +STBIDEF int stbi_is_hdr_from_memory(stbi_uc const *buffer, int len) +{ + #ifndef STBI_NO_HDR + stbi__context s; + stbi__start_mem(&s,buffer,len); + return stbi__hdr_test(&s); + #else + STBI_NOTUSED(buffer); + STBI_NOTUSED(len); + return 0; + #endif +} + +#ifndef STBI_NO_STDIO +STBIDEF int stbi_is_hdr (char const *filename) +{ + FILE *f = stbi__fopen(filename, "rb"); + int result=0; + if (f) { + result = stbi_is_hdr_from_file(f); + fclose(f); + } + return result; +} + +STBIDEF int stbi_is_hdr_from_file(FILE *f) +{ + #ifndef STBI_NO_HDR + long pos = ftell(f); + int res; + stbi__context s; + stbi__start_file(&s,f); + res = stbi__hdr_test(&s); + fseek(f, pos, SEEK_SET); + return res; + #else + STBI_NOTUSED(f); + return 0; + #endif +} +#endif // !STBI_NO_STDIO + +STBIDEF int stbi_is_hdr_from_callbacks(stbi_io_callbacks const *clbk, void *user) +{ + #ifndef STBI_NO_HDR + stbi__context s; + stbi__start_callbacks(&s, (stbi_io_callbacks *) clbk, user); + return stbi__hdr_test(&s); + #else + STBI_NOTUSED(clbk); + STBI_NOTUSED(user); + return 0; + #endif +} + +#ifndef STBI_NO_LINEAR +static float stbi__l2h_gamma=2.2f, stbi__l2h_scale=1.0f; + +STBIDEF void stbi_ldr_to_hdr_gamma(float gamma) { stbi__l2h_gamma = gamma; } +STBIDEF void stbi_ldr_to_hdr_scale(float scale) { stbi__l2h_scale = scale; } +#endif + +static float stbi__h2l_gamma_i=1.0f/2.2f, stbi__h2l_scale_i=1.0f; + +STBIDEF void stbi_hdr_to_ldr_gamma(float gamma) { stbi__h2l_gamma_i = 1/gamma; } +STBIDEF void stbi_hdr_to_ldr_scale(float scale) { stbi__h2l_scale_i = 1/scale; } + + +////////////////////////////////////////////////////////////////////////////// +// +// Common code used by all image loaders +// + +enum +{ + STBI__SCAN_load=0, + STBI__SCAN_type, + STBI__SCAN_header +}; + +static void stbi__refill_buffer(stbi__context *s) +{ + int n = (s->io.read)(s->io_user_data,(char*)s->buffer_start,s->buflen); + if (n == 0) { + // at end of file, treat same as if from memory, but need to handle case + // where s->img_buffer isn't pointing to safe memory, e.g. 0-byte file + s->read_from_callbacks = 0; + s->img_buffer = s->buffer_start; + s->img_buffer_end = s->buffer_start+1; + *s->img_buffer = 0; + } else { + s->img_buffer = s->buffer_start; + s->img_buffer_end = s->buffer_start + n; + } +} + +stbi_inline static stbi_uc stbi__get8(stbi__context *s) +{ + if (s->img_buffer < s->img_buffer_end) + return *s->img_buffer++; + if (s->read_from_callbacks) { + stbi__refill_buffer(s); + return *s->img_buffer++; + } + return 0; +} + +#if defined(STBI_NO_JPEG) && defined(STBI_NO_HDR) && defined(STBI_NO_PIC) && defined(STBI_NO_PNM) +// nothing +#else +stbi_inline static int stbi__at_eof(stbi__context *s) +{ + if (s->io.read) { + if (!(s->io.eof)(s->io_user_data)) return 0; + // if feof() is true, check if buffer = end + // special case: we've only got the special 0 character at the end + if (s->read_from_callbacks == 0) return 1; + } + + return s->img_buffer >= s->img_buffer_end; +} +#endif + +#if defined(STBI_NO_JPEG) && defined(STBI_NO_PNG) && defined(STBI_NO_BMP) && defined(STBI_NO_PSD) && defined(STBI_NO_TGA) && defined(STBI_NO_GIF) && defined(STBI_NO_PIC) +// nothing +#else +static void stbi__skip(stbi__context *s, int n) +{ + if (n < 0) { + s->img_buffer = s->img_buffer_end; + return; + } + if (s->io.read) { + int blen = (int) (s->img_buffer_end - s->img_buffer); + if (blen < n) { + s->img_buffer = s->img_buffer_end; + (s->io.skip)(s->io_user_data, n - blen); + return; + } + } + s->img_buffer += n; +} +#endif + +#if defined(STBI_NO_PNG) && defined(STBI_NO_TGA) && defined(STBI_NO_HDR) && defined(STBI_NO_PNM) +// nothing +#else +static int stbi__getn(stbi__context *s, stbi_uc *buffer, int n) +{ + if (s->io.read) { + int blen = (int) (s->img_buffer_end - s->img_buffer); + if (blen < n) { + int res, count; + + memcpy(buffer, s->img_buffer, blen); + + count = (s->io.read)(s->io_user_data, (char*) buffer + blen, n - blen); + res = (count == (n-blen)); + s->img_buffer = s->img_buffer_end; + return res; + } + } + + if (s->img_buffer+n <= s->img_buffer_end) { + memcpy(buffer, s->img_buffer, n); + s->img_buffer += n; + return 1; + } else + return 0; +} +#endif + +#if defined(STBI_NO_JPEG) && defined(STBI_NO_PNG) && defined(STBI_NO_PSD) && defined(STBI_NO_PIC) +// nothing +#else +static int stbi__get16be(stbi__context *s) +{ + int z = stbi__get8(s); + return (z << 8) + stbi__get8(s); +} +#endif + +#if defined(STBI_NO_PNG) && defined(STBI_NO_PSD) && defined(STBI_NO_PIC) +// nothing +#else +static stbi__uint32 stbi__get32be(stbi__context *s) +{ + stbi__uint32 z = stbi__get16be(s); + return (z << 16) + stbi__get16be(s); +} +#endif + +#if defined(STBI_NO_BMP) && defined(STBI_NO_TGA) && defined(STBI_NO_GIF) +// nothing +#else +static int stbi__get16le(stbi__context *s) +{ + int z = stbi__get8(s); + return z + (stbi__get8(s) << 8); +} +#endif + +#ifndef STBI_NO_BMP +static stbi__uint32 stbi__get32le(stbi__context *s) +{ + stbi__uint32 z = stbi__get16le(s); + return z + (stbi__get16le(s) << 16); +} +#endif + +#define STBI__BYTECAST(x) ((stbi_uc) ((x) & 255)) // truncate int to byte without warnings + +#if defined(STBI_NO_JPEG) && defined(STBI_NO_PNG) && defined(STBI_NO_BMP) && defined(STBI_NO_PSD) && defined(STBI_NO_TGA) && defined(STBI_NO_GIF) && defined(STBI_NO_PIC) && defined(STBI_NO_PNM) +// nothing +#else +////////////////////////////////////////////////////////////////////////////// +// +// generic converter from built-in img_n to req_comp +// individual types do this automatically as much as possible (e.g. jpeg +// does all cases internally since it needs to colorspace convert anyway, +// and it never has alpha, so very few cases ). png can automatically +// interleave an alpha=255 channel, but falls back to this for other cases +// +// assume data buffer is malloced, so malloc a new one and free that one +// only failure mode is malloc failing + +static stbi_uc stbi__compute_y(int r, int g, int b) +{ + return (stbi_uc) (((r*77) + (g*150) + (29*b)) >> 8); +} +#endif + +#if defined(STBI_NO_PNG) && defined(STBI_NO_BMP) && defined(STBI_NO_PSD) && defined(STBI_NO_TGA) && defined(STBI_NO_GIF) && defined(STBI_NO_PIC) && defined(STBI_NO_PNM) +// nothing +#else +static unsigned char *stbi__convert_format(unsigned char *data, int img_n, int req_comp, unsigned int x, unsigned int y) +{ + int i,j; + unsigned char *good; + + if (req_comp == img_n) return data; + STBI_ASSERT(req_comp >= 1 && req_comp <= 4); + + good = (unsigned char *) stbi__malloc_mad3(req_comp, x, y, 0); + if (good == NULL) { + STBI_FREE(data); + return stbi__errpuc("outofmem", "Out of memory"); + } + + for (j=0; j < (int) y; ++j) { + unsigned char *src = data + j * x * img_n ; + unsigned char *dest = good + j * x * req_comp; + + #define STBI__COMBO(a,b) ((a)*8+(b)) + #define STBI__CASE(a,b) case STBI__COMBO(a,b): for(i=x-1; i >= 0; --i, src += a, dest += b) + // convert source image with img_n components to one with req_comp components; + // avoid switch per pixel, so use switch per scanline and massive macros + switch (STBI__COMBO(img_n, req_comp)) { + STBI__CASE(1,2) { dest[0]=src[0]; dest[1]=255; } break; + STBI__CASE(1,3) { dest[0]=dest[1]=dest[2]=src[0]; } break; + STBI__CASE(1,4) { dest[0]=dest[1]=dest[2]=src[0]; dest[3]=255; } break; + STBI__CASE(2,1) { dest[0]=src[0]; } break; + STBI__CASE(2,3) { dest[0]=dest[1]=dest[2]=src[0]; } break; + STBI__CASE(2,4) { dest[0]=dest[1]=dest[2]=src[0]; dest[3]=src[1]; } break; + STBI__CASE(3,4) { dest[0]=src[0];dest[1]=src[1];dest[2]=src[2];dest[3]=255; } break; + STBI__CASE(3,1) { dest[0]=stbi__compute_y(src[0],src[1],src[2]); } break; + STBI__CASE(3,2) { dest[0]=stbi__compute_y(src[0],src[1],src[2]); dest[1] = 255; } break; + STBI__CASE(4,1) { dest[0]=stbi__compute_y(src[0],src[1],src[2]); } break; + STBI__CASE(4,2) { dest[0]=stbi__compute_y(src[0],src[1],src[2]); dest[1] = src[3]; } break; + STBI__CASE(4,3) { dest[0]=src[0];dest[1]=src[1];dest[2]=src[2]; } break; + default: STBI_ASSERT(0); + } + #undef STBI__CASE + } + + STBI_FREE(data); + return good; +} +#endif + +#if defined(STBI_NO_PNG) && defined(STBI_NO_PSD) +// nothing +#else +static stbi__uint16 stbi__compute_y_16(int r, int g, int b) +{ + return (stbi__uint16) (((r*77) + (g*150) + (29*b)) >> 8); +} +#endif + +#if defined(STBI_NO_PNG) && defined(STBI_NO_PSD) +// nothing +#else +static stbi__uint16 *stbi__convert_format16(stbi__uint16 *data, int img_n, int req_comp, unsigned int x, unsigned int y) +{ + int i,j; + stbi__uint16 *good; + + if (req_comp == img_n) return data; + STBI_ASSERT(req_comp >= 1 && req_comp <= 4); + + good = (stbi__uint16 *) stbi__malloc(req_comp * x * y * 2); + if (good == NULL) { + STBI_FREE(data); + return (stbi__uint16 *) stbi__errpuc("outofmem", "Out of memory"); + } + + for (j=0; j < (int) y; ++j) { + stbi__uint16 *src = data + j * x * img_n ; + stbi__uint16 *dest = good + j * x * req_comp; + + #define STBI__COMBO(a,b) ((a)*8+(b)) + #define STBI__CASE(a,b) case STBI__COMBO(a,b): for(i=x-1; i >= 0; --i, src += a, dest += b) + // convert source image with img_n components to one with req_comp components; + // avoid switch per pixel, so use switch per scanline and massive macros + switch (STBI__COMBO(img_n, req_comp)) { + STBI__CASE(1,2) { dest[0]=src[0]; dest[1]=0xffff; } break; + STBI__CASE(1,3) { dest[0]=dest[1]=dest[2]=src[0]; } break; + STBI__CASE(1,4) { dest[0]=dest[1]=dest[2]=src[0]; dest[3]=0xffff; } break; + STBI__CASE(2,1) { dest[0]=src[0]; } break; + STBI__CASE(2,3) { dest[0]=dest[1]=dest[2]=src[0]; } break; + STBI__CASE(2,4) { dest[0]=dest[1]=dest[2]=src[0]; dest[3]=src[1]; } break; + STBI__CASE(3,4) { dest[0]=src[0];dest[1]=src[1];dest[2]=src[2];dest[3]=0xffff; } break; + STBI__CASE(3,1) { dest[0]=stbi__compute_y_16(src[0],src[1],src[2]); } break; + STBI__CASE(3,2) { dest[0]=stbi__compute_y_16(src[0],src[1],src[2]); dest[1] = 0xffff; } break; + STBI__CASE(4,1) { dest[0]=stbi__compute_y_16(src[0],src[1],src[2]); } break; + STBI__CASE(4,2) { dest[0]=stbi__compute_y_16(src[0],src[1],src[2]); dest[1] = src[3]; } break; + STBI__CASE(4,3) { dest[0]=src[0];dest[1]=src[1];dest[2]=src[2]; } break; + default: STBI_ASSERT(0); + } + #undef STBI__CASE + } + + STBI_FREE(data); + return good; +} +#endif + +#ifndef STBI_NO_LINEAR +static float *stbi__ldr_to_hdr(stbi_uc *data, int x, int y, int comp) +{ + int i,k,n; + float *output; + if (!data) return NULL; + output = (float *) stbi__malloc_mad4(x, y, comp, sizeof(float), 0); + if (output == NULL) { STBI_FREE(data); return stbi__errpf("outofmem", "Out of memory"); } + // compute number of non-alpha components + if (comp & 1) n = comp; else n = comp-1; + for (i=0; i < x*y; ++i) { + for (k=0; k < n; ++k) { + output[i*comp + k] = (float) (pow(data[i*comp+k]/255.0f, stbi__l2h_gamma) * stbi__l2h_scale); + } + } + if (n < comp) { + for (i=0; i < x*y; ++i) { + output[i*comp + n] = data[i*comp + n]/255.0f; + } + } + STBI_FREE(data); + return output; +} +#endif + +#ifndef STBI_NO_HDR +#define stbi__float2int(x) ((int) (x)) +static stbi_uc *stbi__hdr_to_ldr(float *data, int x, int y, int comp) +{ + int i,k,n; + stbi_uc *output; + if (!data) return NULL; + output = (stbi_uc *) stbi__malloc_mad3(x, y, comp, 0); + if (output == NULL) { STBI_FREE(data); return stbi__errpuc("outofmem", "Out of memory"); } + // compute number of non-alpha components + if (comp & 1) n = comp; else n = comp-1; + for (i=0; i < x*y; ++i) { + for (k=0; k < n; ++k) { + float z = (float) pow(data[i*comp+k]*stbi__h2l_scale_i, stbi__h2l_gamma_i) * 255 + 0.5f; + if (z < 0) z = 0; + if (z > 255) z = 255; + output[i*comp + k] = (stbi_uc) stbi__float2int(z); + } + if (k < comp) { + float z = data[i*comp+k] * 255 + 0.5f; + if (z < 0) z = 0; + if (z > 255) z = 255; + output[i*comp + k] = (stbi_uc) stbi__float2int(z); + } + } + STBI_FREE(data); + return output; +} +#endif + +////////////////////////////////////////////////////////////////////////////// +// +// "baseline" JPEG/JFIF decoder +// +// simple implementation +// - doesn't support delayed output of y-dimension +// - simple interface (only one output format: 8-bit interleaved RGB) +// - doesn't try to recover corrupt jpegs +// - doesn't allow partial loading, loading multiple at once +// - still fast on x86 (copying globals into locals doesn't help x86) +// - allocates lots of intermediate memory (full size of all components) +// - non-interleaved case requires this anyway +// - allows good upsampling (see next) +// high-quality +// - upsampled channels are bilinearly interpolated, even across blocks +// - quality integer IDCT derived from IJG's 'slow' +// performance +// - fast huffman; reasonable integer IDCT +// - some SIMD kernels for common paths on targets with SSE2/NEON +// - uses a lot of intermediate memory, could cache poorly + +#ifndef STBI_NO_JPEG + +// huffman decoding acceleration +#define FAST_BITS 9 // larger handles more cases; smaller stomps less cache + +typedef struct +{ + stbi_uc fast[1 << FAST_BITS]; + // weirdly, repacking this into AoS is a 10% speed loss, instead of a win + stbi__uint16 code[256]; + stbi_uc values[256]; + stbi_uc size[257]; + unsigned int maxcode[18]; + int delta[17]; // old 'firstsymbol' - old 'firstcode' +} stbi__huffman; + +typedef struct +{ + stbi__context *s; + stbi__huffman huff_dc[4]; + stbi__huffman huff_ac[4]; + stbi__uint16 dequant[4][64]; + stbi__int16 fast_ac[4][1 << FAST_BITS]; + +// sizes for components, interleaved MCUs + int img_h_max, img_v_max; + int img_mcu_x, img_mcu_y; + int img_mcu_w, img_mcu_h; + +// definition of jpeg image component + struct + { + int id; + int h,v; + int tq; + int hd,ha; + int dc_pred; + + int x,y,w2,h2; + stbi_uc *data; + void *raw_data, *raw_coeff; + stbi_uc *linebuf; + short *coeff; // progressive only + int coeff_w, coeff_h; // number of 8x8 coefficient blocks + } img_comp[4]; + + stbi__uint32 code_buffer; // jpeg entropy-coded buffer + int code_bits; // number of valid bits + unsigned char marker; // marker seen while filling entropy buffer + int nomore; // flag if we saw a marker so must stop + + int progressive; + int spec_start; + int spec_end; + int succ_high; + int succ_low; + int eob_run; + int jfif; + int app14_color_transform; // Adobe APP14 tag + int rgb; + + int scan_n, order[4]; + int restart_interval, todo; + +// kernels + void (*idct_block_kernel)(stbi_uc *out, int out_stride, short data[64]); + void (*YCbCr_to_RGB_kernel)(stbi_uc *out, const stbi_uc *y, const stbi_uc *pcb, const stbi_uc *pcr, int count, int step); + stbi_uc *(*resample_row_hv_2_kernel)(stbi_uc *out, stbi_uc *in_near, stbi_uc *in_far, int w, int hs); +} stbi__jpeg; + +static int stbi__build_huffman(stbi__huffman *h, int *count) +{ + int i,j,k=0; + unsigned int code; + // build size list for each symbol (from JPEG spec) + for (i=0; i < 16; ++i) + for (j=0; j < count[i]; ++j) + h->size[k++] = (stbi_uc) (i+1); + h->size[k] = 0; + + // compute actual symbols (from jpeg spec) + code = 0; + k = 0; + for(j=1; j <= 16; ++j) { + // compute delta to add to code to compute symbol id + h->delta[j] = k - code; + if (h->size[k] == j) { + while (h->size[k] == j) + h->code[k++] = (stbi__uint16) (code++); + if (code-1 >= (1u << j)) return stbi__err("bad code lengths","Corrupt JPEG"); + } + // compute largest code + 1 for this size, preshifted as needed later + h->maxcode[j] = code << (16-j); + code <<= 1; + } + h->maxcode[j] = 0xffffffff; + + // build non-spec acceleration table; 255 is flag for not-accelerated + memset(h->fast, 255, 1 << FAST_BITS); + for (i=0; i < k; ++i) { + int s = h->size[i]; + if (s <= FAST_BITS) { + int c = h->code[i] << (FAST_BITS-s); + int m = 1 << (FAST_BITS-s); + for (j=0; j < m; ++j) { + h->fast[c+j] = (stbi_uc) i; + } + } + } + return 1; +} + +// build a table that decodes both magnitude and value of small ACs in +// one go. +static void stbi__build_fast_ac(stbi__int16 *fast_ac, stbi__huffman *h) +{ + int i; + for (i=0; i < (1 << FAST_BITS); ++i) { + stbi_uc fast = h->fast[i]; + fast_ac[i] = 0; + if (fast < 255) { + int rs = h->values[fast]; + int run = (rs >> 4) & 15; + int magbits = rs & 15; + int len = h->size[fast]; + + if (magbits && len + magbits <= FAST_BITS) { + // magnitude code followed by receive_extend code + int k = ((i << len) & ((1 << FAST_BITS) - 1)) >> (FAST_BITS - magbits); + int m = 1 << (magbits - 1); + if (k < m) k += (~0U << magbits) + 1; + // if the result is small enough, we can fit it in fast_ac table + if (k >= -128 && k <= 127) + fast_ac[i] = (stbi__int16) ((k * 256) + (run * 16) + (len + magbits)); + } + } + } +} + +static void stbi__grow_buffer_unsafe(stbi__jpeg *j) +{ + do { + unsigned int b = j->nomore ? 0 : stbi__get8(j->s); + if (b == 0xff) { + int c = stbi__get8(j->s); + while (c == 0xff) c = stbi__get8(j->s); // consume fill bytes + if (c != 0) { + j->marker = (unsigned char) c; + j->nomore = 1; + return; + } + } + j->code_buffer |= b << (24 - j->code_bits); + j->code_bits += 8; + } while (j->code_bits <= 24); +} + +// (1 << n) - 1 +static const stbi__uint32 stbi__bmask[17]={0,1,3,7,15,31,63,127,255,511,1023,2047,4095,8191,16383,32767,65535}; + +// decode a jpeg huffman value from the bitstream +stbi_inline static int stbi__jpeg_huff_decode(stbi__jpeg *j, stbi__huffman *h) +{ + unsigned int temp; + int c,k; + + if (j->code_bits < 16) stbi__grow_buffer_unsafe(j); + + // look at the top FAST_BITS and determine what symbol ID it is, + // if the code is <= FAST_BITS + c = (j->code_buffer >> (32 - FAST_BITS)) & ((1 << FAST_BITS)-1); + k = h->fast[c]; + if (k < 255) { + int s = h->size[k]; + if (s > j->code_bits) + return -1; + j->code_buffer <<= s; + j->code_bits -= s; + return h->values[k]; + } + + // naive test is to shift the code_buffer down so k bits are + // valid, then test against maxcode. To speed this up, we've + // preshifted maxcode left so that it has (16-k) 0s at the + // end; in other words, regardless of the number of bits, it + // wants to be compared against something shifted to have 16; + // that way we don't need to shift inside the loop. + temp = j->code_buffer >> 16; + for (k=FAST_BITS+1 ; ; ++k) + if (temp < h->maxcode[k]) + break; + if (k == 17) { + // error! code not found + j->code_bits -= 16; + return -1; + } + + if (k > j->code_bits) + return -1; + + // convert the huffman code to the symbol id + c = ((j->code_buffer >> (32 - k)) & stbi__bmask[k]) + h->delta[k]; + STBI_ASSERT((((j->code_buffer) >> (32 - h->size[c])) & stbi__bmask[h->size[c]]) == h->code[c]); + + // convert the id to a symbol + j->code_bits -= k; + j->code_buffer <<= k; + return h->values[c]; +} + +// bias[n] = (-1<code_bits < n) stbi__grow_buffer_unsafe(j); + + sgn = (stbi__int32)j->code_buffer >> 31; // sign bit is always in MSB + k = stbi_lrot(j->code_buffer, n); + STBI_ASSERT(n >= 0 && n < (int) (sizeof(stbi__bmask)/sizeof(*stbi__bmask))); + j->code_buffer = k & ~stbi__bmask[n]; + k &= stbi__bmask[n]; + j->code_bits -= n; + return k + (stbi__jbias[n] & ~sgn); +} + +// get some unsigned bits +stbi_inline static int stbi__jpeg_get_bits(stbi__jpeg *j, int n) +{ + unsigned int k; + if (j->code_bits < n) stbi__grow_buffer_unsafe(j); + k = stbi_lrot(j->code_buffer, n); + j->code_buffer = k & ~stbi__bmask[n]; + k &= stbi__bmask[n]; + j->code_bits -= n; + return k; +} + +stbi_inline static int stbi__jpeg_get_bit(stbi__jpeg *j) +{ + unsigned int k; + if (j->code_bits < 1) stbi__grow_buffer_unsafe(j); + k = j->code_buffer; + j->code_buffer <<= 1; + --j->code_bits; + return k & 0x80000000; +} + +// given a value that's at position X in the zigzag stream, +// where does it appear in the 8x8 matrix coded as row-major? +static const stbi_uc stbi__jpeg_dezigzag[64+15] = +{ + 0, 1, 8, 16, 9, 2, 3, 10, + 17, 24, 32, 25, 18, 11, 4, 5, + 12, 19, 26, 33, 40, 48, 41, 34, + 27, 20, 13, 6, 7, 14, 21, 28, + 35, 42, 49, 56, 57, 50, 43, 36, + 29, 22, 15, 23, 30, 37, 44, 51, + 58, 59, 52, 45, 38, 31, 39, 46, + 53, 60, 61, 54, 47, 55, 62, 63, + // let corrupt input sample past end + 63, 63, 63, 63, 63, 63, 63, 63, + 63, 63, 63, 63, 63, 63, 63 +}; + +// decode one 64-entry block-- +static int stbi__jpeg_decode_block(stbi__jpeg *j, short data[64], stbi__huffman *hdc, stbi__huffman *hac, stbi__int16 *fac, int b, stbi__uint16 *dequant) +{ + int diff,dc,k; + int t; + + if (j->code_bits < 16) stbi__grow_buffer_unsafe(j); + t = stbi__jpeg_huff_decode(j, hdc); + if (t < 0) return stbi__err("bad huffman code","Corrupt JPEG"); + + // 0 all the ac values now so we can do it 32-bits at a time + memset(data,0,64*sizeof(data[0])); + + diff = t ? stbi__extend_receive(j, t) : 0; + dc = j->img_comp[b].dc_pred + diff; + j->img_comp[b].dc_pred = dc; + data[0] = (short) (dc * dequant[0]); + + // decode AC components, see JPEG spec + k = 1; + do { + unsigned int zig; + int c,r,s; + if (j->code_bits < 16) stbi__grow_buffer_unsafe(j); + c = (j->code_buffer >> (32 - FAST_BITS)) & ((1 << FAST_BITS)-1); + r = fac[c]; + if (r) { // fast-AC path + k += (r >> 4) & 15; // run + s = r & 15; // combined length + j->code_buffer <<= s; + j->code_bits -= s; + // decode into unzigzag'd location + zig = stbi__jpeg_dezigzag[k++]; + data[zig] = (short) ((r >> 8) * dequant[zig]); + } else { + int rs = stbi__jpeg_huff_decode(j, hac); + if (rs < 0) return stbi__err("bad huffman code","Corrupt JPEG"); + s = rs & 15; + r = rs >> 4; + if (s == 0) { + if (rs != 0xf0) break; // end block + k += 16; + } else { + k += r; + // decode into unzigzag'd location + zig = stbi__jpeg_dezigzag[k++]; + data[zig] = (short) (stbi__extend_receive(j,s) * dequant[zig]); + } + } + } while (k < 64); + return 1; +} + +static int stbi__jpeg_decode_block_prog_dc(stbi__jpeg *j, short data[64], stbi__huffman *hdc, int b) +{ + int diff,dc; + int t; + if (j->spec_end != 0) return stbi__err("can't merge dc and ac", "Corrupt JPEG"); + + if (j->code_bits < 16) stbi__grow_buffer_unsafe(j); + + if (j->succ_high == 0) { + // first scan for DC coefficient, must be first + memset(data,0,64*sizeof(data[0])); // 0 all the ac values now + t = stbi__jpeg_huff_decode(j, hdc); + diff = t ? stbi__extend_receive(j, t) : 0; + + dc = j->img_comp[b].dc_pred + diff; + j->img_comp[b].dc_pred = dc; + data[0] = (short) (dc << j->succ_low); + } else { + // refinement scan for DC coefficient + if (stbi__jpeg_get_bit(j)) + data[0] += (short) (1 << j->succ_low); + } + return 1; +} + +// @OPTIMIZE: store non-zigzagged during the decode passes, +// and only de-zigzag when dequantizing +static int stbi__jpeg_decode_block_prog_ac(stbi__jpeg *j, short data[64], stbi__huffman *hac, stbi__int16 *fac) +{ + int k; + if (j->spec_start == 0) return stbi__err("can't merge dc and ac", "Corrupt JPEG"); + + if (j->succ_high == 0) { + int shift = j->succ_low; + + if (j->eob_run) { + --j->eob_run; + return 1; + } + + k = j->spec_start; + do { + unsigned int zig; + int c,r,s; + if (j->code_bits < 16) stbi__grow_buffer_unsafe(j); + c = (j->code_buffer >> (32 - FAST_BITS)) & ((1 << FAST_BITS)-1); + r = fac[c]; + if (r) { // fast-AC path + k += (r >> 4) & 15; // run + s = r & 15; // combined length + j->code_buffer <<= s; + j->code_bits -= s; + zig = stbi__jpeg_dezigzag[k++]; + data[zig] = (short) ((r >> 8) << shift); + } else { + int rs = stbi__jpeg_huff_decode(j, hac); + if (rs < 0) return stbi__err("bad huffman code","Corrupt JPEG"); + s = rs & 15; + r = rs >> 4; + if (s == 0) { + if (r < 15) { + j->eob_run = (1 << r); + if (r) + j->eob_run += stbi__jpeg_get_bits(j, r); + --j->eob_run; + break; + } + k += 16; + } else { + k += r; + zig = stbi__jpeg_dezigzag[k++]; + data[zig] = (short) (stbi__extend_receive(j,s) << shift); + } + } + } while (k <= j->spec_end); + } else { + // refinement scan for these AC coefficients + + short bit = (short) (1 << j->succ_low); + + if (j->eob_run) { + --j->eob_run; + for (k = j->spec_start; k <= j->spec_end; ++k) { + short *p = &data[stbi__jpeg_dezigzag[k]]; + if (*p != 0) + if (stbi__jpeg_get_bit(j)) + if ((*p & bit)==0) { + if (*p > 0) + *p += bit; + else + *p -= bit; + } + } + } else { + k = j->spec_start; + do { + int r,s; + int rs = stbi__jpeg_huff_decode(j, hac); // @OPTIMIZE see if we can use the fast path here, advance-by-r is so slow, eh + if (rs < 0) return stbi__err("bad huffman code","Corrupt JPEG"); + s = rs & 15; + r = rs >> 4; + if (s == 0) { + if (r < 15) { + j->eob_run = (1 << r) - 1; + if (r) + j->eob_run += stbi__jpeg_get_bits(j, r); + r = 64; // force end of block + } else { + // r=15 s=0 should write 16 0s, so we just do + // a run of 15 0s and then write s (which is 0), + // so we don't have to do anything special here + } + } else { + if (s != 1) return stbi__err("bad huffman code", "Corrupt JPEG"); + // sign bit + if (stbi__jpeg_get_bit(j)) + s = bit; + else + s = -bit; + } + + // advance by r + while (k <= j->spec_end) { + short *p = &data[stbi__jpeg_dezigzag[k++]]; + if (*p != 0) { + if (stbi__jpeg_get_bit(j)) + if ((*p & bit)==0) { + if (*p > 0) + *p += bit; + else + *p -= bit; + } + } else { + if (r == 0) { + *p = (short) s; + break; + } + --r; + } + } + } while (k <= j->spec_end); + } + } + return 1; +} + +// take a -128..127 value and stbi__clamp it and convert to 0..255 +stbi_inline static stbi_uc stbi__clamp(int x) +{ + // trick to use a single test to catch both cases + if ((unsigned int) x > 255) { + if (x < 0) return 0; + if (x > 255) return 255; + } + return (stbi_uc) x; +} + +#define stbi__f2f(x) ((int) (((x) * 4096 + 0.5))) +#define stbi__fsh(x) ((x) * 4096) + +// derived from jidctint -- DCT_ISLOW +#define STBI__IDCT_1D(s0,s1,s2,s3,s4,s5,s6,s7) \ + int t0,t1,t2,t3,p1,p2,p3,p4,p5,x0,x1,x2,x3; \ + p2 = s2; \ + p3 = s6; \ + p1 = (p2+p3) * stbi__f2f(0.5411961f); \ + t2 = p1 + p3*stbi__f2f(-1.847759065f); \ + t3 = p1 + p2*stbi__f2f( 0.765366865f); \ + p2 = s0; \ + p3 = s4; \ + t0 = stbi__fsh(p2+p3); \ + t1 = stbi__fsh(p2-p3); \ + x0 = t0+t3; \ + x3 = t0-t3; \ + x1 = t1+t2; \ + x2 = t1-t2; \ + t0 = s7; \ + t1 = s5; \ + t2 = s3; \ + t3 = s1; \ + p3 = t0+t2; \ + p4 = t1+t3; \ + p1 = t0+t3; \ + p2 = t1+t2; \ + p5 = (p3+p4)*stbi__f2f( 1.175875602f); \ + t0 = t0*stbi__f2f( 0.298631336f); \ + t1 = t1*stbi__f2f( 2.053119869f); \ + t2 = t2*stbi__f2f( 3.072711026f); \ + t3 = t3*stbi__f2f( 1.501321110f); \ + p1 = p5 + p1*stbi__f2f(-0.899976223f); \ + p2 = p5 + p2*stbi__f2f(-2.562915447f); \ + p3 = p3*stbi__f2f(-1.961570560f); \ + p4 = p4*stbi__f2f(-0.390180644f); \ + t3 += p1+p4; \ + t2 += p2+p3; \ + t1 += p2+p4; \ + t0 += p1+p3; + +static void stbi__idct_block(stbi_uc *out, int out_stride, short data[64]) +{ + int i,val[64],*v=val; + stbi_uc *o; + short *d = data; + + // columns + for (i=0; i < 8; ++i,++d, ++v) { + // if all zeroes, shortcut -- this avoids dequantizing 0s and IDCTing + if (d[ 8]==0 && d[16]==0 && d[24]==0 && d[32]==0 + && d[40]==0 && d[48]==0 && d[56]==0) { + // no shortcut 0 seconds + // (1|2|3|4|5|6|7)==0 0 seconds + // all separate -0.047 seconds + // 1 && 2|3 && 4|5 && 6|7: -0.047 seconds + int dcterm = d[0]*4; + v[0] = v[8] = v[16] = v[24] = v[32] = v[40] = v[48] = v[56] = dcterm; + } else { + STBI__IDCT_1D(d[ 0],d[ 8],d[16],d[24],d[32],d[40],d[48],d[56]) + // constants scaled things up by 1<<12; let's bring them back + // down, but keep 2 extra bits of precision + x0 += 512; x1 += 512; x2 += 512; x3 += 512; + v[ 0] = (x0+t3) >> 10; + v[56] = (x0-t3) >> 10; + v[ 8] = (x1+t2) >> 10; + v[48] = (x1-t2) >> 10; + v[16] = (x2+t1) >> 10; + v[40] = (x2-t1) >> 10; + v[24] = (x3+t0) >> 10; + v[32] = (x3-t0) >> 10; + } + } + + for (i=0, v=val, o=out; i < 8; ++i,v+=8,o+=out_stride) { + // no fast case since the first 1D IDCT spread components out + STBI__IDCT_1D(v[0],v[1],v[2],v[3],v[4],v[5],v[6],v[7]) + // constants scaled things up by 1<<12, plus we had 1<<2 from first + // loop, plus horizontal and vertical each scale by sqrt(8) so together + // we've got an extra 1<<3, so 1<<17 total we need to remove. + // so we want to round that, which means adding 0.5 * 1<<17, + // aka 65536. Also, we'll end up with -128 to 127 that we want + // to encode as 0..255 by adding 128, so we'll add that before the shift + x0 += 65536 + (128<<17); + x1 += 65536 + (128<<17); + x2 += 65536 + (128<<17); + x3 += 65536 + (128<<17); + // tried computing the shifts into temps, or'ing the temps to see + // if any were out of range, but that was slower + o[0] = stbi__clamp((x0+t3) >> 17); + o[7] = stbi__clamp((x0-t3) >> 17); + o[1] = stbi__clamp((x1+t2) >> 17); + o[6] = stbi__clamp((x1-t2) >> 17); + o[2] = stbi__clamp((x2+t1) >> 17); + o[5] = stbi__clamp((x2-t1) >> 17); + o[3] = stbi__clamp((x3+t0) >> 17); + o[4] = stbi__clamp((x3-t0) >> 17); + } +} + +#ifdef STBI_SSE2 +// sse2 integer IDCT. not the fastest possible implementation but it +// produces bit-identical results to the generic C version so it's +// fully "transparent". +static void stbi__idct_simd(stbi_uc *out, int out_stride, short data[64]) +{ + // This is constructed to match our regular (generic) integer IDCT exactly. + __m128i row0, row1, row2, row3, row4, row5, row6, row7; + __m128i tmp; + + // dot product constant: even elems=x, odd elems=y + #define dct_const(x,y) _mm_setr_epi16((x),(y),(x),(y),(x),(y),(x),(y)) + + // out(0) = c0[even]*x + c0[odd]*y (c0, x, y 16-bit, out 32-bit) + // out(1) = c1[even]*x + c1[odd]*y + #define dct_rot(out0,out1, x,y,c0,c1) \ + __m128i c0##lo = _mm_unpacklo_epi16((x),(y)); \ + __m128i c0##hi = _mm_unpackhi_epi16((x),(y)); \ + __m128i out0##_l = _mm_madd_epi16(c0##lo, c0); \ + __m128i out0##_h = _mm_madd_epi16(c0##hi, c0); \ + __m128i out1##_l = _mm_madd_epi16(c0##lo, c1); \ + __m128i out1##_h = _mm_madd_epi16(c0##hi, c1) + + // out = in << 12 (in 16-bit, out 32-bit) + #define dct_widen(out, in) \ + __m128i out##_l = _mm_srai_epi32(_mm_unpacklo_epi16(_mm_setzero_si128(), (in)), 4); \ + __m128i out##_h = _mm_srai_epi32(_mm_unpackhi_epi16(_mm_setzero_si128(), (in)), 4) + + // wide add + #define dct_wadd(out, a, b) \ + __m128i out##_l = _mm_add_epi32(a##_l, b##_l); \ + __m128i out##_h = _mm_add_epi32(a##_h, b##_h) + + // wide sub + #define dct_wsub(out, a, b) \ + __m128i out##_l = _mm_sub_epi32(a##_l, b##_l); \ + __m128i out##_h = _mm_sub_epi32(a##_h, b##_h) + + // butterfly a/b, add bias, then shift by "s" and pack + #define dct_bfly32o(out0, out1, a,b,bias,s) \ + { \ + __m128i abiased_l = _mm_add_epi32(a##_l, bias); \ + __m128i abiased_h = _mm_add_epi32(a##_h, bias); \ + dct_wadd(sum, abiased, b); \ + dct_wsub(dif, abiased, b); \ + out0 = _mm_packs_epi32(_mm_srai_epi32(sum_l, s), _mm_srai_epi32(sum_h, s)); \ + out1 = _mm_packs_epi32(_mm_srai_epi32(dif_l, s), _mm_srai_epi32(dif_h, s)); \ + } + + // 8-bit interleave step (for transposes) + #define dct_interleave8(a, b) \ + tmp = a; \ + a = _mm_unpacklo_epi8(a, b); \ + b = _mm_unpackhi_epi8(tmp, b) + + // 16-bit interleave step (for transposes) + #define dct_interleave16(a, b) \ + tmp = a; \ + a = _mm_unpacklo_epi16(a, b); \ + b = _mm_unpackhi_epi16(tmp, b) + + #define dct_pass(bias,shift) \ + { \ + /* even part */ \ + dct_rot(t2e,t3e, row2,row6, rot0_0,rot0_1); \ + __m128i sum04 = _mm_add_epi16(row0, row4); \ + __m128i dif04 = _mm_sub_epi16(row0, row4); \ + dct_widen(t0e, sum04); \ + dct_widen(t1e, dif04); \ + dct_wadd(x0, t0e, t3e); \ + dct_wsub(x3, t0e, t3e); \ + dct_wadd(x1, t1e, t2e); \ + dct_wsub(x2, t1e, t2e); \ + /* odd part */ \ + dct_rot(y0o,y2o, row7,row3, rot2_0,rot2_1); \ + dct_rot(y1o,y3o, row5,row1, rot3_0,rot3_1); \ + __m128i sum17 = _mm_add_epi16(row1, row7); \ + __m128i sum35 = _mm_add_epi16(row3, row5); \ + dct_rot(y4o,y5o, sum17,sum35, rot1_0,rot1_1); \ + dct_wadd(x4, y0o, y4o); \ + dct_wadd(x5, y1o, y5o); \ + dct_wadd(x6, y2o, y5o); \ + dct_wadd(x7, y3o, y4o); \ + dct_bfly32o(row0,row7, x0,x7,bias,shift); \ + dct_bfly32o(row1,row6, x1,x6,bias,shift); \ + dct_bfly32o(row2,row5, x2,x5,bias,shift); \ + dct_bfly32o(row3,row4, x3,x4,bias,shift); \ + } + + __m128i rot0_0 = dct_const(stbi__f2f(0.5411961f), stbi__f2f(0.5411961f) + stbi__f2f(-1.847759065f)); + __m128i rot0_1 = dct_const(stbi__f2f(0.5411961f) + stbi__f2f( 0.765366865f), stbi__f2f(0.5411961f)); + __m128i rot1_0 = dct_const(stbi__f2f(1.175875602f) + stbi__f2f(-0.899976223f), stbi__f2f(1.175875602f)); + __m128i rot1_1 = dct_const(stbi__f2f(1.175875602f), stbi__f2f(1.175875602f) + stbi__f2f(-2.562915447f)); + __m128i rot2_0 = dct_const(stbi__f2f(-1.961570560f) + stbi__f2f( 0.298631336f), stbi__f2f(-1.961570560f)); + __m128i rot2_1 = dct_const(stbi__f2f(-1.961570560f), stbi__f2f(-1.961570560f) + stbi__f2f( 3.072711026f)); + __m128i rot3_0 = dct_const(stbi__f2f(-0.390180644f) + stbi__f2f( 2.053119869f), stbi__f2f(-0.390180644f)); + __m128i rot3_1 = dct_const(stbi__f2f(-0.390180644f), stbi__f2f(-0.390180644f) + stbi__f2f( 1.501321110f)); + + // rounding biases in column/row passes, see stbi__idct_block for explanation. + __m128i bias_0 = _mm_set1_epi32(512); + __m128i bias_1 = _mm_set1_epi32(65536 + (128<<17)); + + // load + row0 = _mm_load_si128((const __m128i *) (data + 0*8)); + row1 = _mm_load_si128((const __m128i *) (data + 1*8)); + row2 = _mm_load_si128((const __m128i *) (data + 2*8)); + row3 = _mm_load_si128((const __m128i *) (data + 3*8)); + row4 = _mm_load_si128((const __m128i *) (data + 4*8)); + row5 = _mm_load_si128((const __m128i *) (data + 5*8)); + row6 = _mm_load_si128((const __m128i *) (data + 6*8)); + row7 = _mm_load_si128((const __m128i *) (data + 7*8)); + + // column pass + dct_pass(bias_0, 10); + + { + // 16bit 8x8 transpose pass 1 + dct_interleave16(row0, row4); + dct_interleave16(row1, row5); + dct_interleave16(row2, row6); + dct_interleave16(row3, row7); + + // transpose pass 2 + dct_interleave16(row0, row2); + dct_interleave16(row1, row3); + dct_interleave16(row4, row6); + dct_interleave16(row5, row7); + + // transpose pass 3 + dct_interleave16(row0, row1); + dct_interleave16(row2, row3); + dct_interleave16(row4, row5); + dct_interleave16(row6, row7); + } + + // row pass + dct_pass(bias_1, 17); + + { + // pack + __m128i p0 = _mm_packus_epi16(row0, row1); // a0a1a2a3...a7b0b1b2b3...b7 + __m128i p1 = _mm_packus_epi16(row2, row3); + __m128i p2 = _mm_packus_epi16(row4, row5); + __m128i p3 = _mm_packus_epi16(row6, row7); + + // 8bit 8x8 transpose pass 1 + dct_interleave8(p0, p2); // a0e0a1e1... + dct_interleave8(p1, p3); // c0g0c1g1... + + // transpose pass 2 + dct_interleave8(p0, p1); // a0c0e0g0... + dct_interleave8(p2, p3); // b0d0f0h0... + + // transpose pass 3 + dct_interleave8(p0, p2); // a0b0c0d0... + dct_interleave8(p1, p3); // a4b4c4d4... + + // store + _mm_storel_epi64((__m128i *) out, p0); out += out_stride; + _mm_storel_epi64((__m128i *) out, _mm_shuffle_epi32(p0, 0x4e)); out += out_stride; + _mm_storel_epi64((__m128i *) out, p2); out += out_stride; + _mm_storel_epi64((__m128i *) out, _mm_shuffle_epi32(p2, 0x4e)); out += out_stride; + _mm_storel_epi64((__m128i *) out, p1); out += out_stride; + _mm_storel_epi64((__m128i *) out, _mm_shuffle_epi32(p1, 0x4e)); out += out_stride; + _mm_storel_epi64((__m128i *) out, p3); out += out_stride; + _mm_storel_epi64((__m128i *) out, _mm_shuffle_epi32(p3, 0x4e)); + } + +#undef dct_const +#undef dct_rot +#undef dct_widen +#undef dct_wadd +#undef dct_wsub +#undef dct_bfly32o +#undef dct_interleave8 +#undef dct_interleave16 +#undef dct_pass +} + +#endif // STBI_SSE2 + +#ifdef STBI_NEON + +// NEON integer IDCT. should produce bit-identical +// results to the generic C version. +static void stbi__idct_simd(stbi_uc *out, int out_stride, short data[64]) +{ + int16x8_t row0, row1, row2, row3, row4, row5, row6, row7; + + int16x4_t rot0_0 = vdup_n_s16(stbi__f2f(0.5411961f)); + int16x4_t rot0_1 = vdup_n_s16(stbi__f2f(-1.847759065f)); + int16x4_t rot0_2 = vdup_n_s16(stbi__f2f( 0.765366865f)); + int16x4_t rot1_0 = vdup_n_s16(stbi__f2f( 1.175875602f)); + int16x4_t rot1_1 = vdup_n_s16(stbi__f2f(-0.899976223f)); + int16x4_t rot1_2 = vdup_n_s16(stbi__f2f(-2.562915447f)); + int16x4_t rot2_0 = vdup_n_s16(stbi__f2f(-1.961570560f)); + int16x4_t rot2_1 = vdup_n_s16(stbi__f2f(-0.390180644f)); + int16x4_t rot3_0 = vdup_n_s16(stbi__f2f( 0.298631336f)); + int16x4_t rot3_1 = vdup_n_s16(stbi__f2f( 2.053119869f)); + int16x4_t rot3_2 = vdup_n_s16(stbi__f2f( 3.072711026f)); + int16x4_t rot3_3 = vdup_n_s16(stbi__f2f( 1.501321110f)); + +#define dct_long_mul(out, inq, coeff) \ + int32x4_t out##_l = vmull_s16(vget_low_s16(inq), coeff); \ + int32x4_t out##_h = vmull_s16(vget_high_s16(inq), coeff) + +#define dct_long_mac(out, acc, inq, coeff) \ + int32x4_t out##_l = vmlal_s16(acc##_l, vget_low_s16(inq), coeff); \ + int32x4_t out##_h = vmlal_s16(acc##_h, vget_high_s16(inq), coeff) + +#define dct_widen(out, inq) \ + int32x4_t out##_l = vshll_n_s16(vget_low_s16(inq), 12); \ + int32x4_t out##_h = vshll_n_s16(vget_high_s16(inq), 12) + +// wide add +#define dct_wadd(out, a, b) \ + int32x4_t out##_l = vaddq_s32(a##_l, b##_l); \ + int32x4_t out##_h = vaddq_s32(a##_h, b##_h) + +// wide sub +#define dct_wsub(out, a, b) \ + int32x4_t out##_l = vsubq_s32(a##_l, b##_l); \ + int32x4_t out##_h = vsubq_s32(a##_h, b##_h) + +// butterfly a/b, then shift using "shiftop" by "s" and pack +#define dct_bfly32o(out0,out1, a,b,shiftop,s) \ + { \ + dct_wadd(sum, a, b); \ + dct_wsub(dif, a, b); \ + out0 = vcombine_s16(shiftop(sum_l, s), shiftop(sum_h, s)); \ + out1 = vcombine_s16(shiftop(dif_l, s), shiftop(dif_h, s)); \ + } + +#define dct_pass(shiftop, shift) \ + { \ + /* even part */ \ + int16x8_t sum26 = vaddq_s16(row2, row6); \ + dct_long_mul(p1e, sum26, rot0_0); \ + dct_long_mac(t2e, p1e, row6, rot0_1); \ + dct_long_mac(t3e, p1e, row2, rot0_2); \ + int16x8_t sum04 = vaddq_s16(row0, row4); \ + int16x8_t dif04 = vsubq_s16(row0, row4); \ + dct_widen(t0e, sum04); \ + dct_widen(t1e, dif04); \ + dct_wadd(x0, t0e, t3e); \ + dct_wsub(x3, t0e, t3e); \ + dct_wadd(x1, t1e, t2e); \ + dct_wsub(x2, t1e, t2e); \ + /* odd part */ \ + int16x8_t sum15 = vaddq_s16(row1, row5); \ + int16x8_t sum17 = vaddq_s16(row1, row7); \ + int16x8_t sum35 = vaddq_s16(row3, row5); \ + int16x8_t sum37 = vaddq_s16(row3, row7); \ + int16x8_t sumodd = vaddq_s16(sum17, sum35); \ + dct_long_mul(p5o, sumodd, rot1_0); \ + dct_long_mac(p1o, p5o, sum17, rot1_1); \ + dct_long_mac(p2o, p5o, sum35, rot1_2); \ + dct_long_mul(p3o, sum37, rot2_0); \ + dct_long_mul(p4o, sum15, rot2_1); \ + dct_wadd(sump13o, p1o, p3o); \ + dct_wadd(sump24o, p2o, p4o); \ + dct_wadd(sump23o, p2o, p3o); \ + dct_wadd(sump14o, p1o, p4o); \ + dct_long_mac(x4, sump13o, row7, rot3_0); \ + dct_long_mac(x5, sump24o, row5, rot3_1); \ + dct_long_mac(x6, sump23o, row3, rot3_2); \ + dct_long_mac(x7, sump14o, row1, rot3_3); \ + dct_bfly32o(row0,row7, x0,x7,shiftop,shift); \ + dct_bfly32o(row1,row6, x1,x6,shiftop,shift); \ + dct_bfly32o(row2,row5, x2,x5,shiftop,shift); \ + dct_bfly32o(row3,row4, x3,x4,shiftop,shift); \ + } + + // load + row0 = vld1q_s16(data + 0*8); + row1 = vld1q_s16(data + 1*8); + row2 = vld1q_s16(data + 2*8); + row3 = vld1q_s16(data + 3*8); + row4 = vld1q_s16(data + 4*8); + row5 = vld1q_s16(data + 5*8); + row6 = vld1q_s16(data + 6*8); + row7 = vld1q_s16(data + 7*8); + + // add DC bias + row0 = vaddq_s16(row0, vsetq_lane_s16(1024, vdupq_n_s16(0), 0)); + + // column pass + dct_pass(vrshrn_n_s32, 10); + + // 16bit 8x8 transpose + { +// these three map to a single VTRN.16, VTRN.32, and VSWP, respectively. +// whether compilers actually get this is another story, sadly. +#define dct_trn16(x, y) { int16x8x2_t t = vtrnq_s16(x, y); x = t.val[0]; y = t.val[1]; } +#define dct_trn32(x, y) { int32x4x2_t t = vtrnq_s32(vreinterpretq_s32_s16(x), vreinterpretq_s32_s16(y)); x = vreinterpretq_s16_s32(t.val[0]); y = vreinterpretq_s16_s32(t.val[1]); } +#define dct_trn64(x, y) { int16x8_t x0 = x; int16x8_t y0 = y; x = vcombine_s16(vget_low_s16(x0), vget_low_s16(y0)); y = vcombine_s16(vget_high_s16(x0), vget_high_s16(y0)); } + + // pass 1 + dct_trn16(row0, row1); // a0b0a2b2a4b4a6b6 + dct_trn16(row2, row3); + dct_trn16(row4, row5); + dct_trn16(row6, row7); + + // pass 2 + dct_trn32(row0, row2); // a0b0c0d0a4b4c4d4 + dct_trn32(row1, row3); + dct_trn32(row4, row6); + dct_trn32(row5, row7); + + // pass 3 + dct_trn64(row0, row4); // a0b0c0d0e0f0g0h0 + dct_trn64(row1, row5); + dct_trn64(row2, row6); + dct_trn64(row3, row7); + +#undef dct_trn16 +#undef dct_trn32 +#undef dct_trn64 + } + + // row pass + // vrshrn_n_s32 only supports shifts up to 16, we need + // 17. so do a non-rounding shift of 16 first then follow + // up with a rounding shift by 1. + dct_pass(vshrn_n_s32, 16); + + { + // pack and round + uint8x8_t p0 = vqrshrun_n_s16(row0, 1); + uint8x8_t p1 = vqrshrun_n_s16(row1, 1); + uint8x8_t p2 = vqrshrun_n_s16(row2, 1); + uint8x8_t p3 = vqrshrun_n_s16(row3, 1); + uint8x8_t p4 = vqrshrun_n_s16(row4, 1); + uint8x8_t p5 = vqrshrun_n_s16(row5, 1); + uint8x8_t p6 = vqrshrun_n_s16(row6, 1); + uint8x8_t p7 = vqrshrun_n_s16(row7, 1); + + // again, these can translate into one instruction, but often don't. +#define dct_trn8_8(x, y) { uint8x8x2_t t = vtrn_u8(x, y); x = t.val[0]; y = t.val[1]; } +#define dct_trn8_16(x, y) { uint16x4x2_t t = vtrn_u16(vreinterpret_u16_u8(x), vreinterpret_u16_u8(y)); x = vreinterpret_u8_u16(t.val[0]); y = vreinterpret_u8_u16(t.val[1]); } +#define dct_trn8_32(x, y) { uint32x2x2_t t = vtrn_u32(vreinterpret_u32_u8(x), vreinterpret_u32_u8(y)); x = vreinterpret_u8_u32(t.val[0]); y = vreinterpret_u8_u32(t.val[1]); } + + // sadly can't use interleaved stores here since we only write + // 8 bytes to each scan line! + + // 8x8 8-bit transpose pass 1 + dct_trn8_8(p0, p1); + dct_trn8_8(p2, p3); + dct_trn8_8(p4, p5); + dct_trn8_8(p6, p7); + + // pass 2 + dct_trn8_16(p0, p2); + dct_trn8_16(p1, p3); + dct_trn8_16(p4, p6); + dct_trn8_16(p5, p7); + + // pass 3 + dct_trn8_32(p0, p4); + dct_trn8_32(p1, p5); + dct_trn8_32(p2, p6); + dct_trn8_32(p3, p7); + + // store + vst1_u8(out, p0); out += out_stride; + vst1_u8(out, p1); out += out_stride; + vst1_u8(out, p2); out += out_stride; + vst1_u8(out, p3); out += out_stride; + vst1_u8(out, p4); out += out_stride; + vst1_u8(out, p5); out += out_stride; + vst1_u8(out, p6); out += out_stride; + vst1_u8(out, p7); + +#undef dct_trn8_8 +#undef dct_trn8_16 +#undef dct_trn8_32 + } + +#undef dct_long_mul +#undef dct_long_mac +#undef dct_widen +#undef dct_wadd +#undef dct_wsub +#undef dct_bfly32o +#undef dct_pass +} + +#endif // STBI_NEON + +#define STBI__MARKER_none 0xff +// if there's a pending marker from the entropy stream, return that +// otherwise, fetch from the stream and get a marker. if there's no +// marker, return 0xff, which is never a valid marker value +static stbi_uc stbi__get_marker(stbi__jpeg *j) +{ + stbi_uc x; + if (j->marker != STBI__MARKER_none) { x = j->marker; j->marker = STBI__MARKER_none; return x; } + x = stbi__get8(j->s); + if (x != 0xff) return STBI__MARKER_none; + while (x == 0xff) + x = stbi__get8(j->s); // consume repeated 0xff fill bytes + return x; +} + +// in each scan, we'll have scan_n components, and the order +// of the components is specified by order[] +#define STBI__RESTART(x) ((x) >= 0xd0 && (x) <= 0xd7) + +// after a restart interval, stbi__jpeg_reset the entropy decoder and +// the dc prediction +static void stbi__jpeg_reset(stbi__jpeg *j) +{ + j->code_bits = 0; + j->code_buffer = 0; + j->nomore = 0; + j->img_comp[0].dc_pred = j->img_comp[1].dc_pred = j->img_comp[2].dc_pred = j->img_comp[3].dc_pred = 0; + j->marker = STBI__MARKER_none; + j->todo = j->restart_interval ? j->restart_interval : 0x7fffffff; + j->eob_run = 0; + // no more than 1<<31 MCUs if no restart_interal? that's plenty safe, + // since we don't even allow 1<<30 pixels +} + +static int stbi__parse_entropy_coded_data(stbi__jpeg *z) +{ + stbi__jpeg_reset(z); + if (!z->progressive) { + if (z->scan_n == 1) { + int i,j; + STBI_SIMD_ALIGN(short, data[64]); + int n = z->order[0]; + // non-interleaved data, we just need to process one block at a time, + // in trivial scanline order + // number of blocks to do just depends on how many actual "pixels" this + // component has, independent of interleaved MCU blocking and such + int w = (z->img_comp[n].x+7) >> 3; + int h = (z->img_comp[n].y+7) >> 3; + for (j=0; j < h; ++j) { + for (i=0; i < w; ++i) { + int ha = z->img_comp[n].ha; + if (!stbi__jpeg_decode_block(z, data, z->huff_dc+z->img_comp[n].hd, z->huff_ac+ha, z->fast_ac[ha], n, z->dequant[z->img_comp[n].tq])) return 0; + z->idct_block_kernel(z->img_comp[n].data+z->img_comp[n].w2*j*8+i*8, z->img_comp[n].w2, data); + // every data block is an MCU, so countdown the restart interval + if (--z->todo <= 0) { + if (z->code_bits < 24) stbi__grow_buffer_unsafe(z); + // if it's NOT a restart, then just bail, so we get corrupt data + // rather than no data + if (!STBI__RESTART(z->marker)) return 1; + stbi__jpeg_reset(z); + } + } + } + return 1; + } else { // interleaved + int i,j,k,x,y; + STBI_SIMD_ALIGN(short, data[64]); + for (j=0; j < z->img_mcu_y; ++j) { + for (i=0; i < z->img_mcu_x; ++i) { + // scan an interleaved mcu... process scan_n components in order + for (k=0; k < z->scan_n; ++k) { + int n = z->order[k]; + // scan out an mcu's worth of this component; that's just determined + // by the basic H and V specified for the component + for (y=0; y < z->img_comp[n].v; ++y) { + for (x=0; x < z->img_comp[n].h; ++x) { + int x2 = (i*z->img_comp[n].h + x)*8; + int y2 = (j*z->img_comp[n].v + y)*8; + int ha = z->img_comp[n].ha; + if (!stbi__jpeg_decode_block(z, data, z->huff_dc+z->img_comp[n].hd, z->huff_ac+ha, z->fast_ac[ha], n, z->dequant[z->img_comp[n].tq])) return 0; + z->idct_block_kernel(z->img_comp[n].data+z->img_comp[n].w2*y2+x2, z->img_comp[n].w2, data); + } + } + } + // after all interleaved components, that's an interleaved MCU, + // so now count down the restart interval + if (--z->todo <= 0) { + if (z->code_bits < 24) stbi__grow_buffer_unsafe(z); + if (!STBI__RESTART(z->marker)) return 1; + stbi__jpeg_reset(z); + } + } + } + return 1; + } + } else { + if (z->scan_n == 1) { + int i,j; + int n = z->order[0]; + // non-interleaved data, we just need to process one block at a time, + // in trivial scanline order + // number of blocks to do just depends on how many actual "pixels" this + // component has, independent of interleaved MCU blocking and such + int w = (z->img_comp[n].x+7) >> 3; + int h = (z->img_comp[n].y+7) >> 3; + for (j=0; j < h; ++j) { + for (i=0; i < w; ++i) { + short *data = z->img_comp[n].coeff + 64 * (i + j * z->img_comp[n].coeff_w); + if (z->spec_start == 0) { + if (!stbi__jpeg_decode_block_prog_dc(z, data, &z->huff_dc[z->img_comp[n].hd], n)) + return 0; + } else { + int ha = z->img_comp[n].ha; + if (!stbi__jpeg_decode_block_prog_ac(z, data, &z->huff_ac[ha], z->fast_ac[ha])) + return 0; + } + // every data block is an MCU, so countdown the restart interval + if (--z->todo <= 0) { + if (z->code_bits < 24) stbi__grow_buffer_unsafe(z); + if (!STBI__RESTART(z->marker)) return 1; + stbi__jpeg_reset(z); + } + } + } + return 1; + } else { // interleaved + int i,j,k,x,y; + for (j=0; j < z->img_mcu_y; ++j) { + for (i=0; i < z->img_mcu_x; ++i) { + // scan an interleaved mcu... process scan_n components in order + for (k=0; k < z->scan_n; ++k) { + int n = z->order[k]; + // scan out an mcu's worth of this component; that's just determined + // by the basic H and V specified for the component + for (y=0; y < z->img_comp[n].v; ++y) { + for (x=0; x < z->img_comp[n].h; ++x) { + int x2 = (i*z->img_comp[n].h + x); + int y2 = (j*z->img_comp[n].v + y); + short *data = z->img_comp[n].coeff + 64 * (x2 + y2 * z->img_comp[n].coeff_w); + if (!stbi__jpeg_decode_block_prog_dc(z, data, &z->huff_dc[z->img_comp[n].hd], n)) + return 0; + } + } + } + // after all interleaved components, that's an interleaved MCU, + // so now count down the restart interval + if (--z->todo <= 0) { + if (z->code_bits < 24) stbi__grow_buffer_unsafe(z); + if (!STBI__RESTART(z->marker)) return 1; + stbi__jpeg_reset(z); + } + } + } + return 1; + } + } +} + +static void stbi__jpeg_dequantize(short *data, stbi__uint16 *dequant) +{ + int i; + for (i=0; i < 64; ++i) + data[i] *= dequant[i]; +} + +static void stbi__jpeg_finish(stbi__jpeg *z) +{ + if (z->progressive) { + // dequantize and idct the data + int i,j,n; + for (n=0; n < z->s->img_n; ++n) { + int w = (z->img_comp[n].x+7) >> 3; + int h = (z->img_comp[n].y+7) >> 3; + for (j=0; j < h; ++j) { + for (i=0; i < w; ++i) { + short *data = z->img_comp[n].coeff + 64 * (i + j * z->img_comp[n].coeff_w); + stbi__jpeg_dequantize(data, z->dequant[z->img_comp[n].tq]); + z->idct_block_kernel(z->img_comp[n].data+z->img_comp[n].w2*j*8+i*8, z->img_comp[n].w2, data); + } + } + } + } +} + +static int stbi__process_marker(stbi__jpeg *z, int m) +{ + int L; + switch (m) { + case STBI__MARKER_none: // no marker found + return stbi__err("expected marker","Corrupt JPEG"); + + case 0xDD: // DRI - specify restart interval + if (stbi__get16be(z->s) != 4) return stbi__err("bad DRI len","Corrupt JPEG"); + z->restart_interval = stbi__get16be(z->s); + return 1; + + case 0xDB: // DQT - define quantization table + L = stbi__get16be(z->s)-2; + while (L > 0) { + int q = stbi__get8(z->s); + int p = q >> 4, sixteen = (p != 0); + int t = q & 15,i; + if (p != 0 && p != 1) return stbi__err("bad DQT type","Corrupt JPEG"); + if (t > 3) return stbi__err("bad DQT table","Corrupt JPEG"); + + for (i=0; i < 64; ++i) + z->dequant[t][stbi__jpeg_dezigzag[i]] = (stbi__uint16)(sixteen ? stbi__get16be(z->s) : stbi__get8(z->s)); + L -= (sixteen ? 129 : 65); + } + return L==0; + + case 0xC4: // DHT - define huffman table + L = stbi__get16be(z->s)-2; + while (L > 0) { + stbi_uc *v; + int sizes[16],i,n=0; + int q = stbi__get8(z->s); + int tc = q >> 4; + int th = q & 15; + if (tc > 1 || th > 3) return stbi__err("bad DHT header","Corrupt JPEG"); + for (i=0; i < 16; ++i) { + sizes[i] = stbi__get8(z->s); + n += sizes[i]; + } + L -= 17; + if (tc == 0) { + if (!stbi__build_huffman(z->huff_dc+th, sizes)) return 0; + v = z->huff_dc[th].values; + } else { + if (!stbi__build_huffman(z->huff_ac+th, sizes)) return 0; + v = z->huff_ac[th].values; + } + for (i=0; i < n; ++i) + v[i] = stbi__get8(z->s); + if (tc != 0) + stbi__build_fast_ac(z->fast_ac[th], z->huff_ac + th); + L -= n; + } + return L==0; + } + + // check for comment block or APP blocks + if ((m >= 0xE0 && m <= 0xEF) || m == 0xFE) { + L = stbi__get16be(z->s); + if (L < 2) { + if (m == 0xFE) + return stbi__err("bad COM len","Corrupt JPEG"); + else + return stbi__err("bad APP len","Corrupt JPEG"); + } + L -= 2; + + if (m == 0xE0 && L >= 5) { // JFIF APP0 segment + static const unsigned char tag[5] = {'J','F','I','F','\0'}; + int ok = 1; + int i; + for (i=0; i < 5; ++i) + if (stbi__get8(z->s) != tag[i]) + ok = 0; + L -= 5; + if (ok) + z->jfif = 1; + } else if (m == 0xEE && L >= 12) { // Adobe APP14 segment + static const unsigned char tag[6] = {'A','d','o','b','e','\0'}; + int ok = 1; + int i; + for (i=0; i < 6; ++i) + if (stbi__get8(z->s) != tag[i]) + ok = 0; + L -= 6; + if (ok) { + stbi__get8(z->s); // version + stbi__get16be(z->s); // flags0 + stbi__get16be(z->s); // flags1 + z->app14_color_transform = stbi__get8(z->s); // color transform + L -= 6; + } + } + + stbi__skip(z->s, L); + return 1; + } + + return stbi__err("unknown marker","Corrupt JPEG"); +} + +// after we see SOS +static int stbi__process_scan_header(stbi__jpeg *z) +{ + int i; + int Ls = stbi__get16be(z->s); + z->scan_n = stbi__get8(z->s); + if (z->scan_n < 1 || z->scan_n > 4 || z->scan_n > (int) z->s->img_n) return stbi__err("bad SOS component count","Corrupt JPEG"); + if (Ls != 6+2*z->scan_n) return stbi__err("bad SOS len","Corrupt JPEG"); + for (i=0; i < z->scan_n; ++i) { + int id = stbi__get8(z->s), which; + int q = stbi__get8(z->s); + for (which = 0; which < z->s->img_n; ++which) + if (z->img_comp[which].id == id) + break; + if (which == z->s->img_n) return 0; // no match + z->img_comp[which].hd = q >> 4; if (z->img_comp[which].hd > 3) return stbi__err("bad DC huff","Corrupt JPEG"); + z->img_comp[which].ha = q & 15; if (z->img_comp[which].ha > 3) return stbi__err("bad AC huff","Corrupt JPEG"); + z->order[i] = which; + } + + { + int aa; + z->spec_start = stbi__get8(z->s); + z->spec_end = stbi__get8(z->s); // should be 63, but might be 0 + aa = stbi__get8(z->s); + z->succ_high = (aa >> 4); + z->succ_low = (aa & 15); + if (z->progressive) { + if (z->spec_start > 63 || z->spec_end > 63 || z->spec_start > z->spec_end || z->succ_high > 13 || z->succ_low > 13) + return stbi__err("bad SOS", "Corrupt JPEG"); + } else { + if (z->spec_start != 0) return stbi__err("bad SOS","Corrupt JPEG"); + if (z->succ_high != 0 || z->succ_low != 0) return stbi__err("bad SOS","Corrupt JPEG"); + z->spec_end = 63; + } + } + + return 1; +} + +static int stbi__free_jpeg_components(stbi__jpeg *z, int ncomp, int why) +{ + int i; + for (i=0; i < ncomp; ++i) { + if (z->img_comp[i].raw_data) { + STBI_FREE(z->img_comp[i].raw_data); + z->img_comp[i].raw_data = NULL; + z->img_comp[i].data = NULL; + } + if (z->img_comp[i].raw_coeff) { + STBI_FREE(z->img_comp[i].raw_coeff); + z->img_comp[i].raw_coeff = 0; + z->img_comp[i].coeff = 0; + } + if (z->img_comp[i].linebuf) { + STBI_FREE(z->img_comp[i].linebuf); + z->img_comp[i].linebuf = NULL; + } + } + return why; +} + +static int stbi__process_frame_header(stbi__jpeg *z, int scan) +{ + stbi__context *s = z->s; + int Lf,p,i,q, h_max=1,v_max=1,c; + Lf = stbi__get16be(s); if (Lf < 11) return stbi__err("bad SOF len","Corrupt JPEG"); // JPEG + p = stbi__get8(s); if (p != 8) return stbi__err("only 8-bit","JPEG format not supported: 8-bit only"); // JPEG baseline + s->img_y = stbi__get16be(s); if (s->img_y == 0) return stbi__err("no header height", "JPEG format not supported: delayed height"); // Legal, but we don't handle it--but neither does IJG + s->img_x = stbi__get16be(s); if (s->img_x == 0) return stbi__err("0 width","Corrupt JPEG"); // JPEG requires + c = stbi__get8(s); + if (c != 3 && c != 1 && c != 4) return stbi__err("bad component count","Corrupt JPEG"); + s->img_n = c; + for (i=0; i < c; ++i) { + z->img_comp[i].data = NULL; + z->img_comp[i].linebuf = NULL; + } + + if (Lf != 8+3*s->img_n) return stbi__err("bad SOF len","Corrupt JPEG"); + + z->rgb = 0; + for (i=0; i < s->img_n; ++i) { + static const unsigned char rgb[3] = { 'R', 'G', 'B' }; + z->img_comp[i].id = stbi__get8(s); + if (s->img_n == 3 && z->img_comp[i].id == rgb[i]) + ++z->rgb; + q = stbi__get8(s); + z->img_comp[i].h = (q >> 4); if (!z->img_comp[i].h || z->img_comp[i].h > 4) return stbi__err("bad H","Corrupt JPEG"); + z->img_comp[i].v = q & 15; if (!z->img_comp[i].v || z->img_comp[i].v > 4) return stbi__err("bad V","Corrupt JPEG"); + z->img_comp[i].tq = stbi__get8(s); if (z->img_comp[i].tq > 3) return stbi__err("bad TQ","Corrupt JPEG"); + } + + if (scan != STBI__SCAN_load) return 1; + + if (!stbi__mad3sizes_valid(s->img_x, s->img_y, s->img_n, 0)) return stbi__err("too large", "Image too large to decode"); + + for (i=0; i < s->img_n; ++i) { + if (z->img_comp[i].h > h_max) h_max = z->img_comp[i].h; + if (z->img_comp[i].v > v_max) v_max = z->img_comp[i].v; + } + + // compute interleaved mcu info + z->img_h_max = h_max; + z->img_v_max = v_max; + z->img_mcu_w = h_max * 8; + z->img_mcu_h = v_max * 8; + // these sizes can't be more than 17 bits + z->img_mcu_x = (s->img_x + z->img_mcu_w-1) / z->img_mcu_w; + z->img_mcu_y = (s->img_y + z->img_mcu_h-1) / z->img_mcu_h; + + for (i=0; i < s->img_n; ++i) { + // number of effective pixels (e.g. for non-interleaved MCU) + z->img_comp[i].x = (s->img_x * z->img_comp[i].h + h_max-1) / h_max; + z->img_comp[i].y = (s->img_y * z->img_comp[i].v + v_max-1) / v_max; + // to simplify generation, we'll allocate enough memory to decode + // the bogus oversized data from using interleaved MCUs and their + // big blocks (e.g. a 16x16 iMCU on an image of width 33); we won't + // discard the extra data until colorspace conversion + // + // img_mcu_x, img_mcu_y: <=17 bits; comp[i].h and .v are <=4 (checked earlier) + // so these muls can't overflow with 32-bit ints (which we require) + z->img_comp[i].w2 = z->img_mcu_x * z->img_comp[i].h * 8; + z->img_comp[i].h2 = z->img_mcu_y * z->img_comp[i].v * 8; + z->img_comp[i].coeff = 0; + z->img_comp[i].raw_coeff = 0; + z->img_comp[i].linebuf = NULL; + z->img_comp[i].raw_data = stbi__malloc_mad2(z->img_comp[i].w2, z->img_comp[i].h2, 15); + if (z->img_comp[i].raw_data == NULL) + return stbi__free_jpeg_components(z, i+1, stbi__err("outofmem", "Out of memory")); + // align blocks for idct using mmx/sse + z->img_comp[i].data = (stbi_uc*) (((size_t) z->img_comp[i].raw_data + 15) & ~15); + if (z->progressive) { + // w2, h2 are multiples of 8 (see above) + z->img_comp[i].coeff_w = z->img_comp[i].w2 / 8; + z->img_comp[i].coeff_h = z->img_comp[i].h2 / 8; + z->img_comp[i].raw_coeff = stbi__malloc_mad3(z->img_comp[i].w2, z->img_comp[i].h2, sizeof(short), 15); + if (z->img_comp[i].raw_coeff == NULL) + return stbi__free_jpeg_components(z, i+1, stbi__err("outofmem", "Out of memory")); + z->img_comp[i].coeff = (short*) (((size_t) z->img_comp[i].raw_coeff + 15) & ~15); + } + } + + return 1; +} + +// use comparisons since in some cases we handle more than one case (e.g. SOF) +#define stbi__DNL(x) ((x) == 0xdc) +#define stbi__SOI(x) ((x) == 0xd8) +#define stbi__EOI(x) ((x) == 0xd9) +#define stbi__SOF(x) ((x) == 0xc0 || (x) == 0xc1 || (x) == 0xc2) +#define stbi__SOS(x) ((x) == 0xda) + +#define stbi__SOF_progressive(x) ((x) == 0xc2) + +static int stbi__decode_jpeg_header(stbi__jpeg *z, int scan) +{ + int m; + z->jfif = 0; + z->app14_color_transform = -1; // valid values are 0,1,2 + z->marker = STBI__MARKER_none; // initialize cached marker to empty + m = stbi__get_marker(z); + if (!stbi__SOI(m)) return stbi__err("no SOI","Corrupt JPEG"); + if (scan == STBI__SCAN_type) return 1; + m = stbi__get_marker(z); + while (!stbi__SOF(m)) { + if (!stbi__process_marker(z,m)) return 0; + m = stbi__get_marker(z); + while (m == STBI__MARKER_none) { + // some files have extra padding after their blocks, so ok, we'll scan + if (stbi__at_eof(z->s)) return stbi__err("no SOF", "Corrupt JPEG"); + m = stbi__get_marker(z); + } + } + z->progressive = stbi__SOF_progressive(m); + if (!stbi__process_frame_header(z, scan)) return 0; + return 1; +} + +// decode image to YCbCr format +static int stbi__decode_jpeg_image(stbi__jpeg *j) +{ + int m; + for (m = 0; m < 4; m++) { + j->img_comp[m].raw_data = NULL; + j->img_comp[m].raw_coeff = NULL; + } + j->restart_interval = 0; + if (!stbi__decode_jpeg_header(j, STBI__SCAN_load)) return 0; + m = stbi__get_marker(j); + while (!stbi__EOI(m)) { + if (stbi__SOS(m)) { + if (!stbi__process_scan_header(j)) return 0; + if (!stbi__parse_entropy_coded_data(j)) return 0; + if (j->marker == STBI__MARKER_none ) { + // handle 0s at the end of image data from IP Kamera 9060 + while (!stbi__at_eof(j->s)) { + int x = stbi__get8(j->s); + if (x == 255) { + j->marker = stbi__get8(j->s); + break; + } + } + // if we reach eof without hitting a marker, stbi__get_marker() below will fail and we'll eventually return 0 + } + } else if (stbi__DNL(m)) { + int Ld = stbi__get16be(j->s); + stbi__uint32 NL = stbi__get16be(j->s); + if (Ld != 4) return stbi__err("bad DNL len", "Corrupt JPEG"); + if (NL != j->s->img_y) return stbi__err("bad DNL height", "Corrupt JPEG"); + } else { + if (!stbi__process_marker(j, m)) return 0; + } + m = stbi__get_marker(j); + } + if (j->progressive) + stbi__jpeg_finish(j); + return 1; +} + +// static jfif-centered resampling (across block boundaries) + +typedef stbi_uc *(*resample_row_func)(stbi_uc *out, stbi_uc *in0, stbi_uc *in1, + int w, int hs); + +#define stbi__div4(x) ((stbi_uc) ((x) >> 2)) + +static stbi_uc *resample_row_1(stbi_uc *out, stbi_uc *in_near, stbi_uc *in_far, int w, int hs) +{ + STBI_NOTUSED(out); + STBI_NOTUSED(in_far); + STBI_NOTUSED(w); + STBI_NOTUSED(hs); + return in_near; +} + +static stbi_uc* stbi__resample_row_v_2(stbi_uc *out, stbi_uc *in_near, stbi_uc *in_far, int w, int hs) +{ + // need to generate two samples vertically for every one in input + int i; + STBI_NOTUSED(hs); + for (i=0; i < w; ++i) + out[i] = stbi__div4(3*in_near[i] + in_far[i] + 2); + return out; +} + +static stbi_uc* stbi__resample_row_h_2(stbi_uc *out, stbi_uc *in_near, stbi_uc *in_far, int w, int hs) +{ + // need to generate two samples horizontally for every one in input + int i; + stbi_uc *input = in_near; + + if (w == 1) { + // if only one sample, can't do any interpolation + out[0] = out[1] = input[0]; + return out; + } + + out[0] = input[0]; + out[1] = stbi__div4(input[0]*3 + input[1] + 2); + for (i=1; i < w-1; ++i) { + int n = 3*input[i]+2; + out[i*2+0] = stbi__div4(n+input[i-1]); + out[i*2+1] = stbi__div4(n+input[i+1]); + } + out[i*2+0] = stbi__div4(input[w-2]*3 + input[w-1] + 2); + out[i*2+1] = input[w-1]; + + STBI_NOTUSED(in_far); + STBI_NOTUSED(hs); + + return out; +} + +#define stbi__div16(x) ((stbi_uc) ((x) >> 4)) + +static stbi_uc *stbi__resample_row_hv_2(stbi_uc *out, stbi_uc *in_near, stbi_uc *in_far, int w, int hs) +{ + // need to generate 2x2 samples for every one in input + int i,t0,t1; + if (w == 1) { + out[0] = out[1] = stbi__div4(3*in_near[0] + in_far[0] + 2); + return out; + } + + t1 = 3*in_near[0] + in_far[0]; + out[0] = stbi__div4(t1+2); + for (i=1; i < w; ++i) { + t0 = t1; + t1 = 3*in_near[i]+in_far[i]; + out[i*2-1] = stbi__div16(3*t0 + t1 + 8); + out[i*2 ] = stbi__div16(3*t1 + t0 + 8); + } + out[w*2-1] = stbi__div4(t1+2); + + STBI_NOTUSED(hs); + + return out; +} + +#if defined(STBI_SSE2) || defined(STBI_NEON) +static stbi_uc *stbi__resample_row_hv_2_simd(stbi_uc *out, stbi_uc *in_near, stbi_uc *in_far, int w, int hs) +{ + // need to generate 2x2 samples for every one in input + int i=0,t0,t1; + + if (w == 1) { + out[0] = out[1] = stbi__div4(3*in_near[0] + in_far[0] + 2); + return out; + } + + t1 = 3*in_near[0] + in_far[0]; + // process groups of 8 pixels for as long as we can. + // note we can't handle the last pixel in a row in this loop + // because we need to handle the filter boundary conditions. + for (; i < ((w-1) & ~7); i += 8) { +#if defined(STBI_SSE2) + // load and perform the vertical filtering pass + // this uses 3*x + y = 4*x + (y - x) + __m128i zero = _mm_setzero_si128(); + __m128i farb = _mm_loadl_epi64((__m128i *) (in_far + i)); + __m128i nearb = _mm_loadl_epi64((__m128i *) (in_near + i)); + __m128i farw = _mm_unpacklo_epi8(farb, zero); + __m128i nearw = _mm_unpacklo_epi8(nearb, zero); + __m128i diff = _mm_sub_epi16(farw, nearw); + __m128i nears = _mm_slli_epi16(nearw, 2); + __m128i curr = _mm_add_epi16(nears, diff); // current row + + // horizontal filter works the same based on shifted vers of current + // row. "prev" is current row shifted right by 1 pixel; we need to + // insert the previous pixel value (from t1). + // "next" is current row shifted left by 1 pixel, with first pixel + // of next block of 8 pixels added in. + __m128i prv0 = _mm_slli_si128(curr, 2); + __m128i nxt0 = _mm_srli_si128(curr, 2); + __m128i prev = _mm_insert_epi16(prv0, t1, 0); + __m128i next = _mm_insert_epi16(nxt0, 3*in_near[i+8] + in_far[i+8], 7); + + // horizontal filter, polyphase implementation since it's convenient: + // even pixels = 3*cur + prev = cur*4 + (prev - cur) + // odd pixels = 3*cur + next = cur*4 + (next - cur) + // note the shared term. + __m128i bias = _mm_set1_epi16(8); + __m128i curs = _mm_slli_epi16(curr, 2); + __m128i prvd = _mm_sub_epi16(prev, curr); + __m128i nxtd = _mm_sub_epi16(next, curr); + __m128i curb = _mm_add_epi16(curs, bias); + __m128i even = _mm_add_epi16(prvd, curb); + __m128i odd = _mm_add_epi16(nxtd, curb); + + // interleave even and odd pixels, then undo scaling. + __m128i int0 = _mm_unpacklo_epi16(even, odd); + __m128i int1 = _mm_unpackhi_epi16(even, odd); + __m128i de0 = _mm_srli_epi16(int0, 4); + __m128i de1 = _mm_srli_epi16(int1, 4); + + // pack and write output + __m128i outv = _mm_packus_epi16(de0, de1); + _mm_storeu_si128((__m128i *) (out + i*2), outv); +#elif defined(STBI_NEON) + // load and perform the vertical filtering pass + // this uses 3*x + y = 4*x + (y - x) + uint8x8_t farb = vld1_u8(in_far + i); + uint8x8_t nearb = vld1_u8(in_near + i); + int16x8_t diff = vreinterpretq_s16_u16(vsubl_u8(farb, nearb)); + int16x8_t nears = vreinterpretq_s16_u16(vshll_n_u8(nearb, 2)); + int16x8_t curr = vaddq_s16(nears, diff); // current row + + // horizontal filter works the same based on shifted vers of current + // row. "prev" is current row shifted right by 1 pixel; we need to + // insert the previous pixel value (from t1). + // "next" is current row shifted left by 1 pixel, with first pixel + // of next block of 8 pixels added in. + int16x8_t prv0 = vextq_s16(curr, curr, 7); + int16x8_t nxt0 = vextq_s16(curr, curr, 1); + int16x8_t prev = vsetq_lane_s16(t1, prv0, 0); + int16x8_t next = vsetq_lane_s16(3*in_near[i+8] + in_far[i+8], nxt0, 7); + + // horizontal filter, polyphase implementation since it's convenient: + // even pixels = 3*cur + prev = cur*4 + (prev - cur) + // odd pixels = 3*cur + next = cur*4 + (next - cur) + // note the shared term. + int16x8_t curs = vshlq_n_s16(curr, 2); + int16x8_t prvd = vsubq_s16(prev, curr); + int16x8_t nxtd = vsubq_s16(next, curr); + int16x8_t even = vaddq_s16(curs, prvd); + int16x8_t odd = vaddq_s16(curs, nxtd); + + // undo scaling and round, then store with even/odd phases interleaved + uint8x8x2_t o; + o.val[0] = vqrshrun_n_s16(even, 4); + o.val[1] = vqrshrun_n_s16(odd, 4); + vst2_u8(out + i*2, o); +#endif + + // "previous" value for next iter + t1 = 3*in_near[i+7] + in_far[i+7]; + } + + t0 = t1; + t1 = 3*in_near[i] + in_far[i]; + out[i*2] = stbi__div16(3*t1 + t0 + 8); + + for (++i; i < w; ++i) { + t0 = t1; + t1 = 3*in_near[i]+in_far[i]; + out[i*2-1] = stbi__div16(3*t0 + t1 + 8); + out[i*2 ] = stbi__div16(3*t1 + t0 + 8); + } + out[w*2-1] = stbi__div4(t1+2); + + STBI_NOTUSED(hs); + + return out; +} +#endif + +static stbi_uc *stbi__resample_row_generic(stbi_uc *out, stbi_uc *in_near, stbi_uc *in_far, int w, int hs) +{ + // resample with nearest-neighbor + int i,j; + STBI_NOTUSED(in_far); + for (i=0; i < w; ++i) + for (j=0; j < hs; ++j) + out[i*hs+j] = in_near[i]; + return out; +} + +// this is a reduced-precision calculation of YCbCr-to-RGB introduced +// to make sure the code produces the same results in both SIMD and scalar +#define stbi__float2fixed(x) (((int) ((x) * 4096.0f + 0.5f)) << 8) +static void stbi__YCbCr_to_RGB_row(stbi_uc *out, const stbi_uc *y, const stbi_uc *pcb, const stbi_uc *pcr, int count, int step) +{ + int i; + for (i=0; i < count; ++i) { + int y_fixed = (y[i] << 20) + (1<<19); // rounding + int r,g,b; + int cr = pcr[i] - 128; + int cb = pcb[i] - 128; + r = y_fixed + cr* stbi__float2fixed(1.40200f); + g = y_fixed + (cr*-stbi__float2fixed(0.71414f)) + ((cb*-stbi__float2fixed(0.34414f)) & 0xffff0000); + b = y_fixed + cb* stbi__float2fixed(1.77200f); + r >>= 20; + g >>= 20; + b >>= 20; + if ((unsigned) r > 255) { if (r < 0) r = 0; else r = 255; } + if ((unsigned) g > 255) { if (g < 0) g = 0; else g = 255; } + if ((unsigned) b > 255) { if (b < 0) b = 0; else b = 255; } + out[0] = (stbi_uc)r; + out[1] = (stbi_uc)g; + out[2] = (stbi_uc)b; + out[3] = 255; + out += step; + } +} + +#if defined(STBI_SSE2) || defined(STBI_NEON) +static void stbi__YCbCr_to_RGB_simd(stbi_uc *out, stbi_uc const *y, stbi_uc const *pcb, stbi_uc const *pcr, int count, int step) +{ + int i = 0; + +#ifdef STBI_SSE2 + // step == 3 is pretty ugly on the final interleave, and i'm not convinced + // it's useful in practice (you wouldn't use it for textures, for example). + // so just accelerate step == 4 case. + if (step == 4) { + // this is a fairly straightforward implementation and not super-optimized. + __m128i signflip = _mm_set1_epi8(-0x80); + __m128i cr_const0 = _mm_set1_epi16( (short) ( 1.40200f*4096.0f+0.5f)); + __m128i cr_const1 = _mm_set1_epi16( - (short) ( 0.71414f*4096.0f+0.5f)); + __m128i cb_const0 = _mm_set1_epi16( - (short) ( 0.34414f*4096.0f+0.5f)); + __m128i cb_const1 = _mm_set1_epi16( (short) ( 1.77200f*4096.0f+0.5f)); + __m128i y_bias = _mm_set1_epi8((char) (unsigned char) 128); + __m128i xw = _mm_set1_epi16(255); // alpha channel + + for (; i+7 < count; i += 8) { + // load + __m128i y_bytes = _mm_loadl_epi64((__m128i *) (y+i)); + __m128i cr_bytes = _mm_loadl_epi64((__m128i *) (pcr+i)); + __m128i cb_bytes = _mm_loadl_epi64((__m128i *) (pcb+i)); + __m128i cr_biased = _mm_xor_si128(cr_bytes, signflip); // -128 + __m128i cb_biased = _mm_xor_si128(cb_bytes, signflip); // -128 + + // unpack to short (and left-shift cr, cb by 8) + __m128i yw = _mm_unpacklo_epi8(y_bias, y_bytes); + __m128i crw = _mm_unpacklo_epi8(_mm_setzero_si128(), cr_biased); + __m128i cbw = _mm_unpacklo_epi8(_mm_setzero_si128(), cb_biased); + + // color transform + __m128i yws = _mm_srli_epi16(yw, 4); + __m128i cr0 = _mm_mulhi_epi16(cr_const0, crw); + __m128i cb0 = _mm_mulhi_epi16(cb_const0, cbw); + __m128i cb1 = _mm_mulhi_epi16(cbw, cb_const1); + __m128i cr1 = _mm_mulhi_epi16(crw, cr_const1); + __m128i rws = _mm_add_epi16(cr0, yws); + __m128i gwt = _mm_add_epi16(cb0, yws); + __m128i bws = _mm_add_epi16(yws, cb1); + __m128i gws = _mm_add_epi16(gwt, cr1); + + // descale + __m128i rw = _mm_srai_epi16(rws, 4); + __m128i bw = _mm_srai_epi16(bws, 4); + __m128i gw = _mm_srai_epi16(gws, 4); + + // back to byte, set up for transpose + __m128i brb = _mm_packus_epi16(rw, bw); + __m128i gxb = _mm_packus_epi16(gw, xw); + + // transpose to interleave channels + __m128i t0 = _mm_unpacklo_epi8(brb, gxb); + __m128i t1 = _mm_unpackhi_epi8(brb, gxb); + __m128i o0 = _mm_unpacklo_epi16(t0, t1); + __m128i o1 = _mm_unpackhi_epi16(t0, t1); + + // store + _mm_storeu_si128((__m128i *) (out + 0), o0); + _mm_storeu_si128((__m128i *) (out + 16), o1); + out += 32; + } + } +#endif + +#ifdef STBI_NEON + // in this version, step=3 support would be easy to add. but is there demand? + if (step == 4) { + // this is a fairly straightforward implementation and not super-optimized. + uint8x8_t signflip = vdup_n_u8(0x80); + int16x8_t cr_const0 = vdupq_n_s16( (short) ( 1.40200f*4096.0f+0.5f)); + int16x8_t cr_const1 = vdupq_n_s16( - (short) ( 0.71414f*4096.0f+0.5f)); + int16x8_t cb_const0 = vdupq_n_s16( - (short) ( 0.34414f*4096.0f+0.5f)); + int16x8_t cb_const1 = vdupq_n_s16( (short) ( 1.77200f*4096.0f+0.5f)); + + for (; i+7 < count; i += 8) { + // load + uint8x8_t y_bytes = vld1_u8(y + i); + uint8x8_t cr_bytes = vld1_u8(pcr + i); + uint8x8_t cb_bytes = vld1_u8(pcb + i); + int8x8_t cr_biased = vreinterpret_s8_u8(vsub_u8(cr_bytes, signflip)); + int8x8_t cb_biased = vreinterpret_s8_u8(vsub_u8(cb_bytes, signflip)); + + // expand to s16 + int16x8_t yws = vreinterpretq_s16_u16(vshll_n_u8(y_bytes, 4)); + int16x8_t crw = vshll_n_s8(cr_biased, 7); + int16x8_t cbw = vshll_n_s8(cb_biased, 7); + + // color transform + int16x8_t cr0 = vqdmulhq_s16(crw, cr_const0); + int16x8_t cb0 = vqdmulhq_s16(cbw, cb_const0); + int16x8_t cr1 = vqdmulhq_s16(crw, cr_const1); + int16x8_t cb1 = vqdmulhq_s16(cbw, cb_const1); + int16x8_t rws = vaddq_s16(yws, cr0); + int16x8_t gws = vaddq_s16(vaddq_s16(yws, cb0), cr1); + int16x8_t bws = vaddq_s16(yws, cb1); + + // undo scaling, round, convert to byte + uint8x8x4_t o; + o.val[0] = vqrshrun_n_s16(rws, 4); + o.val[1] = vqrshrun_n_s16(gws, 4); + o.val[2] = vqrshrun_n_s16(bws, 4); + o.val[3] = vdup_n_u8(255); + + // store, interleaving r/g/b/a + vst4_u8(out, o); + out += 8*4; + } + } +#endif + + for (; i < count; ++i) { + int y_fixed = (y[i] << 20) + (1<<19); // rounding + int r,g,b; + int cr = pcr[i] - 128; + int cb = pcb[i] - 128; + r = y_fixed + cr* stbi__float2fixed(1.40200f); + g = y_fixed + cr*-stbi__float2fixed(0.71414f) + ((cb*-stbi__float2fixed(0.34414f)) & 0xffff0000); + b = y_fixed + cb* stbi__float2fixed(1.77200f); + r >>= 20; + g >>= 20; + b >>= 20; + if ((unsigned) r > 255) { if (r < 0) r = 0; else r = 255; } + if ((unsigned) g > 255) { if (g < 0) g = 0; else g = 255; } + if ((unsigned) b > 255) { if (b < 0) b = 0; else b = 255; } + out[0] = (stbi_uc)r; + out[1] = (stbi_uc)g; + out[2] = (stbi_uc)b; + out[3] = 255; + out += step; + } +} +#endif + +// set up the kernels +static void stbi__setup_jpeg(stbi__jpeg *j) +{ + j->idct_block_kernel = stbi__idct_block; + j->YCbCr_to_RGB_kernel = stbi__YCbCr_to_RGB_row; + j->resample_row_hv_2_kernel = stbi__resample_row_hv_2; + +#ifdef STBI_SSE2 + if (stbi__sse2_available()) { + j->idct_block_kernel = stbi__idct_simd; + j->YCbCr_to_RGB_kernel = stbi__YCbCr_to_RGB_simd; + j->resample_row_hv_2_kernel = stbi__resample_row_hv_2_simd; + } +#endif + +#ifdef STBI_NEON + j->idct_block_kernel = stbi__idct_simd; + j->YCbCr_to_RGB_kernel = stbi__YCbCr_to_RGB_simd; + j->resample_row_hv_2_kernel = stbi__resample_row_hv_2_simd; +#endif +} + +// clean up the temporary component buffers +static void stbi__cleanup_jpeg(stbi__jpeg *j) +{ + stbi__free_jpeg_components(j, j->s->img_n, 0); +} + +typedef struct +{ + resample_row_func resample; + stbi_uc *line0,*line1; + int hs,vs; // expansion factor in each axis + int w_lores; // horizontal pixels pre-expansion + int ystep; // how far through vertical expansion we are + int ypos; // which pre-expansion row we're on +} stbi__resample; + +// fast 0..255 * 0..255 => 0..255 rounded multiplication +static stbi_uc stbi__blinn_8x8(stbi_uc x, stbi_uc y) +{ + unsigned int t = x*y + 128; + return (stbi_uc) ((t + (t >>8)) >> 8); +} + +static stbi_uc *load_jpeg_image(stbi__jpeg *z, int *out_x, int *out_y, int *comp, int req_comp) +{ + int n, decode_n, is_rgb; + z->s->img_n = 0; // make stbi__cleanup_jpeg safe + + // validate req_comp + if (req_comp < 0 || req_comp > 4) return stbi__errpuc("bad req_comp", "Internal error"); + + // load a jpeg image from whichever source, but leave in YCbCr format + if (!stbi__decode_jpeg_image(z)) { stbi__cleanup_jpeg(z); return NULL; } + + // determine actual number of components to generate + n = req_comp ? req_comp : z->s->img_n >= 3 ? 3 : 1; + + is_rgb = z->s->img_n == 3 && (z->rgb == 3 || (z->app14_color_transform == 0 && !z->jfif)); + + if (z->s->img_n == 3 && n < 3 && !is_rgb) + decode_n = 1; + else + decode_n = z->s->img_n; + + // resample and color-convert + { + int k; + unsigned int i,j; + stbi_uc *output; + stbi_uc *coutput[4] = { NULL, NULL, NULL, NULL }; + + stbi__resample res_comp[4]; + + for (k=0; k < decode_n; ++k) { + stbi__resample *r = &res_comp[k]; + + // allocate line buffer big enough for upsampling off the edges + // with upsample factor of 4 + z->img_comp[k].linebuf = (stbi_uc *) stbi__malloc(z->s->img_x + 3); + if (!z->img_comp[k].linebuf) { stbi__cleanup_jpeg(z); return stbi__errpuc("outofmem", "Out of memory"); } + + r->hs = z->img_h_max / z->img_comp[k].h; + r->vs = z->img_v_max / z->img_comp[k].v; + r->ystep = r->vs >> 1; + r->w_lores = (z->s->img_x + r->hs-1) / r->hs; + r->ypos = 0; + r->line0 = r->line1 = z->img_comp[k].data; + + if (r->hs == 1 && r->vs == 1) r->resample = resample_row_1; + else if (r->hs == 1 && r->vs == 2) r->resample = stbi__resample_row_v_2; + else if (r->hs == 2 && r->vs == 1) r->resample = stbi__resample_row_h_2; + else if (r->hs == 2 && r->vs == 2) r->resample = z->resample_row_hv_2_kernel; + else r->resample = stbi__resample_row_generic; + } + + // can't error after this so, this is safe + output = (stbi_uc *) stbi__malloc_mad3(n, z->s->img_x, z->s->img_y, 1); + if (!output) { stbi__cleanup_jpeg(z); return stbi__errpuc("outofmem", "Out of memory"); } + + // now go ahead and resample + for (j=0; j < z->s->img_y; ++j) { + stbi_uc *out = output + n * z->s->img_x * j; + for (k=0; k < decode_n; ++k) { + stbi__resample *r = &res_comp[k]; + int y_bot = r->ystep >= (r->vs >> 1); + coutput[k] = r->resample(z->img_comp[k].linebuf, + y_bot ? r->line1 : r->line0, + y_bot ? r->line0 : r->line1, + r->w_lores, r->hs); + if (++r->ystep >= r->vs) { + r->ystep = 0; + r->line0 = r->line1; + if (++r->ypos < z->img_comp[k].y) + r->line1 += z->img_comp[k].w2; + } + } + if (n >= 3) { + stbi_uc *y = coutput[0]; + if (z->s->img_n == 3) { + if (is_rgb) { + for (i=0; i < z->s->img_x; ++i) { + out[0] = y[i]; + out[1] = coutput[1][i]; + out[2] = coutput[2][i]; + out[3] = 255; + out += n; + } + } else { + z->YCbCr_to_RGB_kernel(out, y, coutput[1], coutput[2], z->s->img_x, n); + } + } else if (z->s->img_n == 4) { + if (z->app14_color_transform == 0) { // CMYK + for (i=0; i < z->s->img_x; ++i) { + stbi_uc m = coutput[3][i]; + out[0] = stbi__blinn_8x8(coutput[0][i], m); + out[1] = stbi__blinn_8x8(coutput[1][i], m); + out[2] = stbi__blinn_8x8(coutput[2][i], m); + out[3] = 255; + out += n; + } + } else if (z->app14_color_transform == 2) { // YCCK + z->YCbCr_to_RGB_kernel(out, y, coutput[1], coutput[2], z->s->img_x, n); + for (i=0; i < z->s->img_x; ++i) { + stbi_uc m = coutput[3][i]; + out[0] = stbi__blinn_8x8(255 - out[0], m); + out[1] = stbi__blinn_8x8(255 - out[1], m); + out[2] = stbi__blinn_8x8(255 - out[2], m); + out += n; + } + } else { // YCbCr + alpha? Ignore the fourth channel for now + z->YCbCr_to_RGB_kernel(out, y, coutput[1], coutput[2], z->s->img_x, n); + } + } else + for (i=0; i < z->s->img_x; ++i) { + out[0] = out[1] = out[2] = y[i]; + out[3] = 255; // not used if n==3 + out += n; + } + } else { + if (is_rgb) { + if (n == 1) + for (i=0; i < z->s->img_x; ++i) + *out++ = stbi__compute_y(coutput[0][i], coutput[1][i], coutput[2][i]); + else { + for (i=0; i < z->s->img_x; ++i, out += 2) { + out[0] = stbi__compute_y(coutput[0][i], coutput[1][i], coutput[2][i]); + out[1] = 255; + } + } + } else if (z->s->img_n == 4 && z->app14_color_transform == 0) { + for (i=0; i < z->s->img_x; ++i) { + stbi_uc m = coutput[3][i]; + stbi_uc r = stbi__blinn_8x8(coutput[0][i], m); + stbi_uc g = stbi__blinn_8x8(coutput[1][i], m); + stbi_uc b = stbi__blinn_8x8(coutput[2][i], m); + out[0] = stbi__compute_y(r, g, b); + out[1] = 255; + out += n; + } + } else if (z->s->img_n == 4 && z->app14_color_transform == 2) { + for (i=0; i < z->s->img_x; ++i) { + out[0] = stbi__blinn_8x8(255 - coutput[0][i], coutput[3][i]); + out[1] = 255; + out += n; + } + } else { + stbi_uc *y = coutput[0]; + if (n == 1) + for (i=0; i < z->s->img_x; ++i) out[i] = y[i]; + else + for (i=0; i < z->s->img_x; ++i) { *out++ = y[i]; *out++ = 255; } + } + } + } + stbi__cleanup_jpeg(z); + *out_x = z->s->img_x; + *out_y = z->s->img_y; + if (comp) *comp = z->s->img_n >= 3 ? 3 : 1; // report original components, not output + return output; + } +} + +static void *stbi__jpeg_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri) +{ + unsigned char* result; + stbi__jpeg* j = (stbi__jpeg*) stbi__malloc(sizeof(stbi__jpeg)); + STBI_NOTUSED(ri); + j->s = s; + stbi__setup_jpeg(j); + result = load_jpeg_image(j, x,y,comp,req_comp); + STBI_FREE(j); + return result; +} + +static int stbi__jpeg_test(stbi__context *s) +{ + int r; + stbi__jpeg* j = (stbi__jpeg*)stbi__malloc(sizeof(stbi__jpeg)); + j->s = s; + stbi__setup_jpeg(j); + r = stbi__decode_jpeg_header(j, STBI__SCAN_type); + stbi__rewind(s); + STBI_FREE(j); + return r; +} + +static int stbi__jpeg_info_raw(stbi__jpeg *j, int *x, int *y, int *comp) +{ + if (!stbi__decode_jpeg_header(j, STBI__SCAN_header)) { + stbi__rewind( j->s ); + return 0; + } + if (x) *x = j->s->img_x; + if (y) *y = j->s->img_y; + if (comp) *comp = j->s->img_n >= 3 ? 3 : 1; + return 1; +} + +static int stbi__jpeg_info(stbi__context *s, int *x, int *y, int *comp) +{ + int result; + stbi__jpeg* j = (stbi__jpeg*) (stbi__malloc(sizeof(stbi__jpeg))); + j->s = s; + result = stbi__jpeg_info_raw(j, x, y, comp); + STBI_FREE(j); + return result; +} +#endif + +// public domain zlib decode v0.2 Sean Barrett 2006-11-18 +// simple implementation +// - all input must be provided in an upfront buffer +// - all output is written to a single output buffer (can malloc/realloc) +// performance +// - fast huffman + +#ifndef STBI_NO_ZLIB + +// fast-way is faster to check than jpeg huffman, but slow way is slower +#define STBI__ZFAST_BITS 9 // accelerate all cases in default tables +#define STBI__ZFAST_MASK ((1 << STBI__ZFAST_BITS) - 1) + +// zlib-style huffman encoding +// (jpegs packs from left, zlib from right, so can't share code) +typedef struct +{ + stbi__uint16 fast[1 << STBI__ZFAST_BITS]; + stbi__uint16 firstcode[16]; + int maxcode[17]; + stbi__uint16 firstsymbol[16]; + stbi_uc size[288]; + stbi__uint16 value[288]; +} stbi__zhuffman; + +stbi_inline static int stbi__bitreverse16(int n) +{ + n = ((n & 0xAAAA) >> 1) | ((n & 0x5555) << 1); + n = ((n & 0xCCCC) >> 2) | ((n & 0x3333) << 2); + n = ((n & 0xF0F0) >> 4) | ((n & 0x0F0F) << 4); + n = ((n & 0xFF00) >> 8) | ((n & 0x00FF) << 8); + return n; +} + +stbi_inline static int stbi__bit_reverse(int v, int bits) +{ + STBI_ASSERT(bits <= 16); + // to bit reverse n bits, reverse 16 and shift + // e.g. 11 bits, bit reverse and shift away 5 + return stbi__bitreverse16(v) >> (16-bits); +} + +static int stbi__zbuild_huffman(stbi__zhuffman *z, const stbi_uc *sizelist, int num) +{ + int i,k=0; + int code, next_code[16], sizes[17]; + + // DEFLATE spec for generating codes + memset(sizes, 0, sizeof(sizes)); + memset(z->fast, 0, sizeof(z->fast)); + for (i=0; i < num; ++i) + ++sizes[sizelist[i]]; + sizes[0] = 0; + for (i=1; i < 16; ++i) + if (sizes[i] > (1 << i)) + return stbi__err("bad sizes", "Corrupt PNG"); + code = 0; + for (i=1; i < 16; ++i) { + next_code[i] = code; + z->firstcode[i] = (stbi__uint16) code; + z->firstsymbol[i] = (stbi__uint16) k; + code = (code + sizes[i]); + if (sizes[i]) + if (code-1 >= (1 << i)) return stbi__err("bad codelengths","Corrupt PNG"); + z->maxcode[i] = code << (16-i); // preshift for inner loop + code <<= 1; + k += sizes[i]; + } + z->maxcode[16] = 0x10000; // sentinel + for (i=0; i < num; ++i) { + int s = sizelist[i]; + if (s) { + int c = next_code[s] - z->firstcode[s] + z->firstsymbol[s]; + stbi__uint16 fastv = (stbi__uint16) ((s << 9) | i); + z->size [c] = (stbi_uc ) s; + z->value[c] = (stbi__uint16) i; + if (s <= STBI__ZFAST_BITS) { + int j = stbi__bit_reverse(next_code[s],s); + while (j < (1 << STBI__ZFAST_BITS)) { + z->fast[j] = fastv; + j += (1 << s); + } + } + ++next_code[s]; + } + } + return 1; +} + +// zlib-from-memory implementation for PNG reading +// because PNG allows splitting the zlib stream arbitrarily, +// and it's annoying structurally to have PNG call ZLIB call PNG, +// we require PNG read all the IDATs and combine them into a single +// memory buffer + +typedef struct +{ + stbi_uc *zbuffer, *zbuffer_end; + int num_bits; + stbi__uint32 code_buffer; + + char *zout; + char *zout_start; + char *zout_end; + int z_expandable; + + stbi__zhuffman z_length, z_distance; +} stbi__zbuf; + +stbi_inline static stbi_uc stbi__zget8(stbi__zbuf *z) +{ + if (z->zbuffer >= z->zbuffer_end) return 0; + return *z->zbuffer++; +} + +static void stbi__fill_bits(stbi__zbuf *z) +{ + do { + STBI_ASSERT(z->code_buffer < (1U << z->num_bits)); + z->code_buffer |= (unsigned int) stbi__zget8(z) << z->num_bits; + z->num_bits += 8; + } while (z->num_bits <= 24); +} + +stbi_inline static unsigned int stbi__zreceive(stbi__zbuf *z, int n) +{ + unsigned int k; + if (z->num_bits < n) stbi__fill_bits(z); + k = z->code_buffer & ((1 << n) - 1); + z->code_buffer >>= n; + z->num_bits -= n; + return k; +} + +static int stbi__zhuffman_decode_slowpath(stbi__zbuf *a, stbi__zhuffman *z) +{ + int b,s,k; + // not resolved by fast table, so compute it the slow way + // use jpeg approach, which requires MSbits at top + k = stbi__bit_reverse(a->code_buffer, 16); + for (s=STBI__ZFAST_BITS+1; ; ++s) + if (k < z->maxcode[s]) + break; + if (s == 16) return -1; // invalid code! + // code size is s, so: + b = (k >> (16-s)) - z->firstcode[s] + z->firstsymbol[s]; + STBI_ASSERT(z->size[b] == s); + a->code_buffer >>= s; + a->num_bits -= s; + return z->value[b]; +} + +stbi_inline static int stbi__zhuffman_decode(stbi__zbuf *a, stbi__zhuffman *z) +{ + int b,s; + if (a->num_bits < 16) stbi__fill_bits(a); + b = z->fast[a->code_buffer & STBI__ZFAST_MASK]; + if (b) { + s = b >> 9; + a->code_buffer >>= s; + a->num_bits -= s; + return b & 511; + } + return stbi__zhuffman_decode_slowpath(a, z); +} + +static int stbi__zexpand(stbi__zbuf *z, char *zout, int n) // need to make room for n bytes +{ + char *q; + int cur, limit, old_limit; + z->zout = zout; + if (!z->z_expandable) return stbi__err("output buffer limit","Corrupt PNG"); + cur = (int) (z->zout - z->zout_start); + limit = old_limit = (int) (z->zout_end - z->zout_start); + while (cur + n > limit) + limit *= 2; + q = (char *) STBI_REALLOC_SIZED(z->zout_start, old_limit, limit); + STBI_NOTUSED(old_limit); + if (q == NULL) return stbi__err("outofmem", "Out of memory"); + z->zout_start = q; + z->zout = q + cur; + z->zout_end = q + limit; + return 1; +} + +static const int stbi__zlength_base[31] = { + 3,4,5,6,7,8,9,10,11,13, + 15,17,19,23,27,31,35,43,51,59, + 67,83,99,115,131,163,195,227,258,0,0 }; + +static const int stbi__zlength_extra[31]= +{ 0,0,0,0,0,0,0,0,1,1,1,1,2,2,2,2,3,3,3,3,4,4,4,4,5,5,5,5,0,0,0 }; + +static const int stbi__zdist_base[32] = { 1,2,3,4,5,7,9,13,17,25,33,49,65,97,129,193, +257,385,513,769,1025,1537,2049,3073,4097,6145,8193,12289,16385,24577,0,0}; + +static const int stbi__zdist_extra[32] = +{ 0,0,0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7,8,8,9,9,10,10,11,11,12,12,13,13}; + +static int stbi__parse_huffman_block(stbi__zbuf *a) +{ + char *zout = a->zout; + for(;;) { + int z = stbi__zhuffman_decode(a, &a->z_length); + if (z < 256) { + if (z < 0) return stbi__err("bad huffman code","Corrupt PNG"); // error in huffman codes + if (zout >= a->zout_end) { + if (!stbi__zexpand(a, zout, 1)) return 0; + zout = a->zout; + } + *zout++ = (char) z; + } else { + stbi_uc *p; + int len,dist; + if (z == 256) { + a->zout = zout; + return 1; + } + z -= 257; + len = stbi__zlength_base[z]; + if (stbi__zlength_extra[z]) len += stbi__zreceive(a, stbi__zlength_extra[z]); + z = stbi__zhuffman_decode(a, &a->z_distance); + if (z < 0) return stbi__err("bad huffman code","Corrupt PNG"); + dist = stbi__zdist_base[z]; + if (stbi__zdist_extra[z]) dist += stbi__zreceive(a, stbi__zdist_extra[z]); + if (zout - a->zout_start < dist) return stbi__err("bad dist","Corrupt PNG"); + if (zout + len > a->zout_end) { + if (!stbi__zexpand(a, zout, len)) return 0; + zout = a->zout; + } + p = (stbi_uc *) (zout - dist); + if (dist == 1) { // run of one byte; common in images. + stbi_uc v = *p; + if (len) { do *zout++ = v; while (--len); } + } else { + if (len) { do *zout++ = *p++; while (--len); } + } + } + } +} + +static int stbi__compute_huffman_codes(stbi__zbuf *a) +{ + static const stbi_uc length_dezigzag[19] = { 16,17,18,0,8,7,9,6,10,5,11,4,12,3,13,2,14,1,15 }; + stbi__zhuffman z_codelength; + stbi_uc lencodes[286+32+137];//padding for maximum single op + stbi_uc codelength_sizes[19]; + int i,n; + + int hlit = stbi__zreceive(a,5) + 257; + int hdist = stbi__zreceive(a,5) + 1; + int hclen = stbi__zreceive(a,4) + 4; + int ntot = hlit + hdist; + + memset(codelength_sizes, 0, sizeof(codelength_sizes)); + for (i=0; i < hclen; ++i) { + int s = stbi__zreceive(a,3); + codelength_sizes[length_dezigzag[i]] = (stbi_uc) s; + } + if (!stbi__zbuild_huffman(&z_codelength, codelength_sizes, 19)) return 0; + + n = 0; + while (n < ntot) { + int c = stbi__zhuffman_decode(a, &z_codelength); + if (c < 0 || c >= 19) return stbi__err("bad codelengths", "Corrupt PNG"); + if (c < 16) + lencodes[n++] = (stbi_uc) c; + else { + stbi_uc fill = 0; + if (c == 16) { + c = stbi__zreceive(a,2)+3; + if (n == 0) return stbi__err("bad codelengths", "Corrupt PNG"); + fill = lencodes[n-1]; + } else if (c == 17) + c = stbi__zreceive(a,3)+3; + else { + STBI_ASSERT(c == 18); + c = stbi__zreceive(a,7)+11; + } + if (ntot - n < c) return stbi__err("bad codelengths", "Corrupt PNG"); + memset(lencodes+n, fill, c); + n += c; + } + } + if (n != ntot) return stbi__err("bad codelengths","Corrupt PNG"); + if (!stbi__zbuild_huffman(&a->z_length, lencodes, hlit)) return 0; + if (!stbi__zbuild_huffman(&a->z_distance, lencodes+hlit, hdist)) return 0; + return 1; +} + +static int stbi__parse_uncompressed_block(stbi__zbuf *a) +{ + stbi_uc header[4]; + int len,nlen,k; + if (a->num_bits & 7) + stbi__zreceive(a, a->num_bits & 7); // discard + // drain the bit-packed data into header + k = 0; + while (a->num_bits > 0) { + header[k++] = (stbi_uc) (a->code_buffer & 255); // suppress MSVC run-time check + a->code_buffer >>= 8; + a->num_bits -= 8; + } + STBI_ASSERT(a->num_bits == 0); + // now fill header the normal way + while (k < 4) + header[k++] = stbi__zget8(a); + len = header[1] * 256 + header[0]; + nlen = header[3] * 256 + header[2]; + if (nlen != (len ^ 0xffff)) return stbi__err("zlib corrupt","Corrupt PNG"); + if (a->zbuffer + len > a->zbuffer_end) return stbi__err("read past buffer","Corrupt PNG"); + if (a->zout + len > a->zout_end) + if (!stbi__zexpand(a, a->zout, len)) return 0; + memcpy(a->zout, a->zbuffer, len); + a->zbuffer += len; + a->zout += len; + return 1; +} + +static int stbi__parse_zlib_header(stbi__zbuf *a) +{ + int cmf = stbi__zget8(a); + int cm = cmf & 15; + /* int cinfo = cmf >> 4; */ + int flg = stbi__zget8(a); + if ((cmf*256+flg) % 31 != 0) return stbi__err("bad zlib header","Corrupt PNG"); // zlib spec + if (flg & 32) return stbi__err("no preset dict","Corrupt PNG"); // preset dictionary not allowed in png + if (cm != 8) return stbi__err("bad compression","Corrupt PNG"); // DEFLATE required for png + // window = 1 << (8 + cinfo)... but who cares, we fully buffer output + return 1; +} + +static const stbi_uc stbi__zdefault_length[288] = +{ + 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, + 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, + 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, + 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, + 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, + 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, + 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, + 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, + 7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7, 7,7,7,7,7,7,7,7,8,8,8,8,8,8,8,8 +}; +static const stbi_uc stbi__zdefault_distance[32] = +{ + 5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5 +}; +/* +Init algorithm: +{ + int i; // use <= to match clearly with spec + for (i=0; i <= 143; ++i) stbi__zdefault_length[i] = 8; + for ( ; i <= 255; ++i) stbi__zdefault_length[i] = 9; + for ( ; i <= 279; ++i) stbi__zdefault_length[i] = 7; + for ( ; i <= 287; ++i) stbi__zdefault_length[i] = 8; + + for (i=0; i <= 31; ++i) stbi__zdefault_distance[i] = 5; +} +*/ + +static int stbi__parse_zlib(stbi__zbuf *a, int parse_header) +{ + int final, type; + if (parse_header) + if (!stbi__parse_zlib_header(a)) return 0; + a->num_bits = 0; + a->code_buffer = 0; + do { + final = stbi__zreceive(a,1); + type = stbi__zreceive(a,2); + if (type == 0) { + if (!stbi__parse_uncompressed_block(a)) return 0; + } else if (type == 3) { + return 0; + } else { + if (type == 1) { + // use fixed code lengths + if (!stbi__zbuild_huffman(&a->z_length , stbi__zdefault_length , 288)) return 0; + if (!stbi__zbuild_huffman(&a->z_distance, stbi__zdefault_distance, 32)) return 0; + } else { + if (!stbi__compute_huffman_codes(a)) return 0; + } + if (!stbi__parse_huffman_block(a)) return 0; + } + } while (!final); + return 1; +} + +static int stbi__do_zlib(stbi__zbuf *a, char *obuf, int olen, int exp, int parse_header) +{ + a->zout_start = obuf; + a->zout = obuf; + a->zout_end = obuf + olen; + a->z_expandable = exp; + + return stbi__parse_zlib(a, parse_header); +} + +STBIDEF char *stbi_zlib_decode_malloc_guesssize(const char *buffer, int len, int initial_size, int *outlen) +{ + stbi__zbuf a; + char *p = (char *) stbi__malloc(initial_size); + if (p == NULL) return NULL; + a.zbuffer = (stbi_uc *) buffer; + a.zbuffer_end = (stbi_uc *) buffer + len; + if (stbi__do_zlib(&a, p, initial_size, 1, 1)) { + if (outlen) *outlen = (int) (a.zout - a.zout_start); + return a.zout_start; + } else { + STBI_FREE(a.zout_start); + return NULL; + } +} + +STBIDEF char *stbi_zlib_decode_malloc(char const *buffer, int len, int *outlen) +{ + return stbi_zlib_decode_malloc_guesssize(buffer, len, 16384, outlen); +} + +STBIDEF char *stbi_zlib_decode_malloc_guesssize_headerflag(const char *buffer, int len, int initial_size, int *outlen, int parse_header) +{ + stbi__zbuf a; + char *p = (char *) stbi__malloc(initial_size); + if (p == NULL) return NULL; + a.zbuffer = (stbi_uc *) buffer; + a.zbuffer_end = (stbi_uc *) buffer + len; + if (stbi__do_zlib(&a, p, initial_size, 1, parse_header)) { + if (outlen) *outlen = (int) (a.zout - a.zout_start); + return a.zout_start; + } else { + STBI_FREE(a.zout_start); + return NULL; + } +} + +STBIDEF int stbi_zlib_decode_buffer(char *obuffer, int olen, char const *ibuffer, int ilen) +{ + stbi__zbuf a; + a.zbuffer = (stbi_uc *) ibuffer; + a.zbuffer_end = (stbi_uc *) ibuffer + ilen; + if (stbi__do_zlib(&a, obuffer, olen, 0, 1)) + return (int) (a.zout - a.zout_start); + else + return -1; +} + +STBIDEF char *stbi_zlib_decode_noheader_malloc(char const *buffer, int len, int *outlen) +{ + stbi__zbuf a; + char *p = (char *) stbi__malloc(16384); + if (p == NULL) return NULL; + a.zbuffer = (stbi_uc *) buffer; + a.zbuffer_end = (stbi_uc *) buffer+len; + if (stbi__do_zlib(&a, p, 16384, 1, 0)) { + if (outlen) *outlen = (int) (a.zout - a.zout_start); + return a.zout_start; + } else { + STBI_FREE(a.zout_start); + return NULL; + } +} + +STBIDEF int stbi_zlib_decode_noheader_buffer(char *obuffer, int olen, const char *ibuffer, int ilen) +{ + stbi__zbuf a; + a.zbuffer = (stbi_uc *) ibuffer; + a.zbuffer_end = (stbi_uc *) ibuffer + ilen; + if (stbi__do_zlib(&a, obuffer, olen, 0, 0)) + return (int) (a.zout - a.zout_start); + else + return -1; +} +#endif + +// public domain "baseline" PNG decoder v0.10 Sean Barrett 2006-11-18 +// simple implementation +// - only 8-bit samples +// - no CRC checking +// - allocates lots of intermediate memory +// - avoids problem of streaming data between subsystems +// - avoids explicit window management +// performance +// - uses stb_zlib, a PD zlib implementation with fast huffman decoding + +#ifndef STBI_NO_PNG +typedef struct +{ + stbi__uint32 length; + stbi__uint32 type; +} stbi__pngchunk; + +static stbi__pngchunk stbi__get_chunk_header(stbi__context *s) +{ + stbi__pngchunk c; + c.length = stbi__get32be(s); + c.type = stbi__get32be(s); + return c; +} + +static int stbi__check_png_header(stbi__context *s) +{ + static const stbi_uc png_sig[8] = { 137,80,78,71,13,10,26,10 }; + int i; + for (i=0; i < 8; ++i) + if (stbi__get8(s) != png_sig[i]) return stbi__err("bad png sig","Not a PNG"); + return 1; +} + +typedef struct +{ + stbi__context *s; + stbi_uc *idata, *expanded, *out; + int depth; +} stbi__png; + + +enum { + STBI__F_none=0, + STBI__F_sub=1, + STBI__F_up=2, + STBI__F_avg=3, + STBI__F_paeth=4, + // synthetic filters used for first scanline to avoid needing a dummy row of 0s + STBI__F_avg_first, + STBI__F_paeth_first +}; + +static stbi_uc first_row_filter[5] = +{ + STBI__F_none, + STBI__F_sub, + STBI__F_none, + STBI__F_avg_first, + STBI__F_paeth_first +}; + +static int stbi__paeth(int a, int b, int c) +{ + int p = a + b - c; + int pa = abs(p-a); + int pb = abs(p-b); + int pc = abs(p-c); + if (pa <= pb && pa <= pc) return a; + if (pb <= pc) return b; + return c; +} + +static const stbi_uc stbi__depth_scale_table[9] = { 0, 0xff, 0x55, 0, 0x11, 0,0,0, 0x01 }; + +// create the png data from post-deflated data +static int stbi__create_png_image_raw(stbi__png *a, stbi_uc *raw, stbi__uint32 raw_len, int out_n, stbi__uint32 x, stbi__uint32 y, int depth, int color) +{ + int bytes = (depth == 16? 2 : 1); + stbi__context *s = a->s; + stbi__uint32 i,j,stride = x*out_n*bytes; + stbi__uint32 img_len, img_width_bytes; + int k; + int img_n = s->img_n; // copy it into a local for later + + int output_bytes = out_n*bytes; + int filter_bytes = img_n*bytes; + int width = x; + + STBI_ASSERT(out_n == s->img_n || out_n == s->img_n+1); + a->out = (stbi_uc *) stbi__malloc_mad3(x, y, output_bytes, 0); // extra bytes to write off the end into + if (!a->out) return stbi__err("outofmem", "Out of memory"); + + if (!stbi__mad3sizes_valid(img_n, x, depth, 7)) return stbi__err("too large", "Corrupt PNG"); + img_width_bytes = (((img_n * x * depth) + 7) >> 3); + img_len = (img_width_bytes + 1) * y; + + // we used to check for exact match between raw_len and img_len on non-interlaced PNGs, + // but issue #276 reported a PNG in the wild that had extra data at the end (all zeros), + // so just check for raw_len < img_len always. + if (raw_len < img_len) return stbi__err("not enough pixels","Corrupt PNG"); + + for (j=0; j < y; ++j) { + stbi_uc *cur = a->out + stride*j; + stbi_uc *prior; + int filter = *raw++; + + if (filter > 4) + return stbi__err("invalid filter","Corrupt PNG"); + + if (depth < 8) { + STBI_ASSERT(img_width_bytes <= x); + cur += x*out_n - img_width_bytes; // store output to the rightmost img_len bytes, so we can decode in place + filter_bytes = 1; + width = img_width_bytes; + } + prior = cur - stride; // bugfix: need to compute this after 'cur +=' computation above + + // if first row, use special filter that doesn't sample previous row + if (j == 0) filter = first_row_filter[filter]; + + // handle first byte explicitly + for (k=0; k < filter_bytes; ++k) { + switch (filter) { + case STBI__F_none : cur[k] = raw[k]; break; + case STBI__F_sub : cur[k] = raw[k]; break; + case STBI__F_up : cur[k] = STBI__BYTECAST(raw[k] + prior[k]); break; + case STBI__F_avg : cur[k] = STBI__BYTECAST(raw[k] + (prior[k]>>1)); break; + case STBI__F_paeth : cur[k] = STBI__BYTECAST(raw[k] + stbi__paeth(0,prior[k],0)); break; + case STBI__F_avg_first : cur[k] = raw[k]; break; + case STBI__F_paeth_first: cur[k] = raw[k]; break; + } + } + + if (depth == 8) { + if (img_n != out_n) + cur[img_n] = 255; // first pixel + raw += img_n; + cur += out_n; + prior += out_n; + } else if (depth == 16) { + if (img_n != out_n) { + cur[filter_bytes] = 255; // first pixel top byte + cur[filter_bytes+1] = 255; // first pixel bottom byte + } + raw += filter_bytes; + cur += output_bytes; + prior += output_bytes; + } else { + raw += 1; + cur += 1; + prior += 1; + } + + // this is a little gross, so that we don't switch per-pixel or per-component + if (depth < 8 || img_n == out_n) { + int nk = (width - 1)*filter_bytes; + #define STBI__CASE(f) \ + case f: \ + for (k=0; k < nk; ++k) + switch (filter) { + // "none" filter turns into a memcpy here; make that explicit. + case STBI__F_none: memcpy(cur, raw, nk); break; + STBI__CASE(STBI__F_sub) { cur[k] = STBI__BYTECAST(raw[k] + cur[k-filter_bytes]); } break; + STBI__CASE(STBI__F_up) { cur[k] = STBI__BYTECAST(raw[k] + prior[k]); } break; + STBI__CASE(STBI__F_avg) { cur[k] = STBI__BYTECAST(raw[k] + ((prior[k] + cur[k-filter_bytes])>>1)); } break; + STBI__CASE(STBI__F_paeth) { cur[k] = STBI__BYTECAST(raw[k] + stbi__paeth(cur[k-filter_bytes],prior[k],prior[k-filter_bytes])); } break; + STBI__CASE(STBI__F_avg_first) { cur[k] = STBI__BYTECAST(raw[k] + (cur[k-filter_bytes] >> 1)); } break; + STBI__CASE(STBI__F_paeth_first) { cur[k] = STBI__BYTECAST(raw[k] + stbi__paeth(cur[k-filter_bytes],0,0)); } break; + } + #undef STBI__CASE + raw += nk; + } else { + STBI_ASSERT(img_n+1 == out_n); + #define STBI__CASE(f) \ + case f: \ + for (i=x-1; i >= 1; --i, cur[filter_bytes]=255,raw+=filter_bytes,cur+=output_bytes,prior+=output_bytes) \ + for (k=0; k < filter_bytes; ++k) + switch (filter) { + STBI__CASE(STBI__F_none) { cur[k] = raw[k]; } break; + STBI__CASE(STBI__F_sub) { cur[k] = STBI__BYTECAST(raw[k] + cur[k- output_bytes]); } break; + STBI__CASE(STBI__F_up) { cur[k] = STBI__BYTECAST(raw[k] + prior[k]); } break; + STBI__CASE(STBI__F_avg) { cur[k] = STBI__BYTECAST(raw[k] + ((prior[k] + cur[k- output_bytes])>>1)); } break; + STBI__CASE(STBI__F_paeth) { cur[k] = STBI__BYTECAST(raw[k] + stbi__paeth(cur[k- output_bytes],prior[k],prior[k- output_bytes])); } break; + STBI__CASE(STBI__F_avg_first) { cur[k] = STBI__BYTECAST(raw[k] + (cur[k- output_bytes] >> 1)); } break; + STBI__CASE(STBI__F_paeth_first) { cur[k] = STBI__BYTECAST(raw[k] + stbi__paeth(cur[k- output_bytes],0,0)); } break; + } + #undef STBI__CASE + + // the loop above sets the high byte of the pixels' alpha, but for + // 16 bit png files we also need the low byte set. we'll do that here. + if (depth == 16) { + cur = a->out + stride*j; // start at the beginning of the row again + for (i=0; i < x; ++i,cur+=output_bytes) { + cur[filter_bytes+1] = 255; + } + } + } + } + + // we make a separate pass to expand bits to pixels; for performance, + // this could run two scanlines behind the above code, so it won't + // intefere with filtering but will still be in the cache. + if (depth < 8) { + for (j=0; j < y; ++j) { + stbi_uc *cur = a->out + stride*j; + stbi_uc *in = a->out + stride*j + x*out_n - img_width_bytes; + // unpack 1/2/4-bit into a 8-bit buffer. allows us to keep the common 8-bit path optimal at minimal cost for 1/2/4-bit + // png guarante byte alignment, if width is not multiple of 8/4/2 we'll decode dummy trailing data that will be skipped in the later loop + stbi_uc scale = (color == 0) ? stbi__depth_scale_table[depth] : 1; // scale grayscale values to 0..255 range + + // note that the final byte might overshoot and write more data than desired. + // we can allocate enough data that this never writes out of memory, but it + // could also overwrite the next scanline. can it overwrite non-empty data + // on the next scanline? yes, consider 1-pixel-wide scanlines with 1-bit-per-pixel. + // so we need to explicitly clamp the final ones + + if (depth == 4) { + for (k=x*img_n; k >= 2; k-=2, ++in) { + *cur++ = scale * ((*in >> 4) ); + *cur++ = scale * ((*in ) & 0x0f); + } + if (k > 0) *cur++ = scale * ((*in >> 4) ); + } else if (depth == 2) { + for (k=x*img_n; k >= 4; k-=4, ++in) { + *cur++ = scale * ((*in >> 6) ); + *cur++ = scale * ((*in >> 4) & 0x03); + *cur++ = scale * ((*in >> 2) & 0x03); + *cur++ = scale * ((*in ) & 0x03); + } + if (k > 0) *cur++ = scale * ((*in >> 6) ); + if (k > 1) *cur++ = scale * ((*in >> 4) & 0x03); + if (k > 2) *cur++ = scale * ((*in >> 2) & 0x03); + } else if (depth == 1) { + for (k=x*img_n; k >= 8; k-=8, ++in) { + *cur++ = scale * ((*in >> 7) ); + *cur++ = scale * ((*in >> 6) & 0x01); + *cur++ = scale * ((*in >> 5) & 0x01); + *cur++ = scale * ((*in >> 4) & 0x01); + *cur++ = scale * ((*in >> 3) & 0x01); + *cur++ = scale * ((*in >> 2) & 0x01); + *cur++ = scale * ((*in >> 1) & 0x01); + *cur++ = scale * ((*in ) & 0x01); + } + if (k > 0) *cur++ = scale * ((*in >> 7) ); + if (k > 1) *cur++ = scale * ((*in >> 6) & 0x01); + if (k > 2) *cur++ = scale * ((*in >> 5) & 0x01); + if (k > 3) *cur++ = scale * ((*in >> 4) & 0x01); + if (k > 4) *cur++ = scale * ((*in >> 3) & 0x01); + if (k > 5) *cur++ = scale * ((*in >> 2) & 0x01); + if (k > 6) *cur++ = scale * ((*in >> 1) & 0x01); + } + if (img_n != out_n) { + int q; + // insert alpha = 255 + cur = a->out + stride*j; + if (img_n == 1) { + for (q=x-1; q >= 0; --q) { + cur[q*2+1] = 255; + cur[q*2+0] = cur[q]; + } + } else { + STBI_ASSERT(img_n == 3); + for (q=x-1; q >= 0; --q) { + cur[q*4+3] = 255; + cur[q*4+2] = cur[q*3+2]; + cur[q*4+1] = cur[q*3+1]; + cur[q*4+0] = cur[q*3+0]; + } + } + } + } + } else if (depth == 16) { + // force the image data from big-endian to platform-native. + // this is done in a separate pass due to the decoding relying + // on the data being untouched, but could probably be done + // per-line during decode if care is taken. + stbi_uc *cur = a->out; + stbi__uint16 *cur16 = (stbi__uint16*)cur; + + for(i=0; i < x*y*out_n; ++i,cur16++,cur+=2) { + *cur16 = (cur[0] << 8) | cur[1]; + } + } + + return 1; +} + +static int stbi__create_png_image(stbi__png *a, stbi_uc *image_data, stbi__uint32 image_data_len, int out_n, int depth, int color, int interlaced) +{ + int bytes = (depth == 16 ? 2 : 1); + int out_bytes = out_n * bytes; + stbi_uc *final; + int p; + if (!interlaced) + return stbi__create_png_image_raw(a, image_data, image_data_len, out_n, a->s->img_x, a->s->img_y, depth, color); + + // de-interlacing + final = (stbi_uc *) stbi__malloc_mad3(a->s->img_x, a->s->img_y, out_bytes, 0); + for (p=0; p < 7; ++p) { + int xorig[] = { 0,4,0,2,0,1,0 }; + int yorig[] = { 0,0,4,0,2,0,1 }; + int xspc[] = { 8,8,4,4,2,2,1 }; + int yspc[] = { 8,8,8,4,4,2,2 }; + int i,j,x,y; + // pass1_x[4] = 0, pass1_x[5] = 1, pass1_x[12] = 1 + x = (a->s->img_x - xorig[p] + xspc[p]-1) / xspc[p]; + y = (a->s->img_y - yorig[p] + yspc[p]-1) / yspc[p]; + if (x && y) { + stbi__uint32 img_len = ((((a->s->img_n * x * depth) + 7) >> 3) + 1) * y; + if (!stbi__create_png_image_raw(a, image_data, image_data_len, out_n, x, y, depth, color)) { + STBI_FREE(final); + return 0; + } + for (j=0; j < y; ++j) { + for (i=0; i < x; ++i) { + int out_y = j*yspc[p]+yorig[p]; + int out_x = i*xspc[p]+xorig[p]; + memcpy(final + out_y*a->s->img_x*out_bytes + out_x*out_bytes, + a->out + (j*x+i)*out_bytes, out_bytes); + } + } + STBI_FREE(a->out); + image_data += img_len; + image_data_len -= img_len; + } + } + a->out = final; + + return 1; +} + +static int stbi__compute_transparency(stbi__png *z, stbi_uc tc[3], int out_n) +{ + stbi__context *s = z->s; + stbi__uint32 i, pixel_count = s->img_x * s->img_y; + stbi_uc *p = z->out; + + // compute color-based transparency, assuming we've + // already got 255 as the alpha value in the output + STBI_ASSERT(out_n == 2 || out_n == 4); + + if (out_n == 2) { + for (i=0; i < pixel_count; ++i) { + p[1] = (p[0] == tc[0] ? 0 : 255); + p += 2; + } + } else { + for (i=0; i < pixel_count; ++i) { + if (p[0] == tc[0] && p[1] == tc[1] && p[2] == tc[2]) + p[3] = 0; + p += 4; + } + } + return 1; +} + +static int stbi__compute_transparency16(stbi__png *z, stbi__uint16 tc[3], int out_n) +{ + stbi__context *s = z->s; + stbi__uint32 i, pixel_count = s->img_x * s->img_y; + stbi__uint16 *p = (stbi__uint16*) z->out; + + // compute color-based transparency, assuming we've + // already got 65535 as the alpha value in the output + STBI_ASSERT(out_n == 2 || out_n == 4); + + if (out_n == 2) { + for (i = 0; i < pixel_count; ++i) { + p[1] = (p[0] == tc[0] ? 0 : 65535); + p += 2; + } + } else { + for (i = 0; i < pixel_count; ++i) { + if (p[0] == tc[0] && p[1] == tc[1] && p[2] == tc[2]) + p[3] = 0; + p += 4; + } + } + return 1; +} + +static int stbi__expand_png_palette(stbi__png *a, stbi_uc *palette, int len, int pal_img_n) +{ + stbi__uint32 i, pixel_count = a->s->img_x * a->s->img_y; + stbi_uc *p, *temp_out, *orig = a->out; + + p = (stbi_uc *) stbi__malloc_mad2(pixel_count, pal_img_n, 0); + if (p == NULL) return stbi__err("outofmem", "Out of memory"); + + // between here and free(out) below, exitting would leak + temp_out = p; + + if (pal_img_n == 3) { + for (i=0; i < pixel_count; ++i) { + int n = orig[i]*4; + p[0] = palette[n ]; + p[1] = palette[n+1]; + p[2] = palette[n+2]; + p += 3; + } + } else { + for (i=0; i < pixel_count; ++i) { + int n = orig[i]*4; + p[0] = palette[n ]; + p[1] = palette[n+1]; + p[2] = palette[n+2]; + p[3] = palette[n+3]; + p += 4; + } + } + STBI_FREE(a->out); + a->out = temp_out; + + STBI_NOTUSED(len); + + return 1; +} + +static int stbi__unpremultiply_on_load = 0; +static int stbi__de_iphone_flag = 0; + +STBIDEF void stbi_set_unpremultiply_on_load(int flag_true_if_should_unpremultiply) +{ + stbi__unpremultiply_on_load = flag_true_if_should_unpremultiply; +} + +STBIDEF void stbi_convert_iphone_png_to_rgb(int flag_true_if_should_convert) +{ + stbi__de_iphone_flag = flag_true_if_should_convert; +} + +static void stbi__de_iphone(stbi__png *z) +{ + stbi__context *s = z->s; + stbi__uint32 i, pixel_count = s->img_x * s->img_y; + stbi_uc *p = z->out; + + if (s->img_out_n == 3) { // convert bgr to rgb + for (i=0; i < pixel_count; ++i) { + stbi_uc t = p[0]; + p[0] = p[2]; + p[2] = t; + p += 3; + } + } else { + STBI_ASSERT(s->img_out_n == 4); + if (stbi__unpremultiply_on_load) { + // convert bgr to rgb and unpremultiply + for (i=0; i < pixel_count; ++i) { + stbi_uc a = p[3]; + stbi_uc t = p[0]; + if (a) { + stbi_uc half = a / 2; + p[0] = (p[2] * 255 + half) / a; + p[1] = (p[1] * 255 + half) / a; + p[2] = ( t * 255 + half) / a; + } else { + p[0] = p[2]; + p[2] = t; + } + p += 4; + } + } else { + // convert bgr to rgb + for (i=0; i < pixel_count; ++i) { + stbi_uc t = p[0]; + p[0] = p[2]; + p[2] = t; + p += 4; + } + } + } +} + +#define STBI__PNG_TYPE(a,b,c,d) (((unsigned) (a) << 24) + ((unsigned) (b) << 16) + ((unsigned) (c) << 8) + (unsigned) (d)) + +static int stbi__parse_png_file(stbi__png *z, int scan, int req_comp) +{ + stbi_uc palette[1024], pal_img_n=0; + stbi_uc has_trans=0, tc[3]={0}; + stbi__uint16 tc16[3]; + stbi__uint32 ioff=0, idata_limit=0, i, pal_len=0; + int first=1,k,interlace=0, color=0, is_iphone=0; + stbi__context *s = z->s; + + z->expanded = NULL; + z->idata = NULL; + z->out = NULL; + + if (!stbi__check_png_header(s)) return 0; + + if (scan == STBI__SCAN_type) return 1; + + for (;;) { + stbi__pngchunk c = stbi__get_chunk_header(s); + switch (c.type) { + case STBI__PNG_TYPE('C','g','B','I'): + is_iphone = 1; + stbi__skip(s, c.length); + break; + case STBI__PNG_TYPE('I','H','D','R'): { + int comp,filter; + if (!first) return stbi__err("multiple IHDR","Corrupt PNG"); + first = 0; + if (c.length != 13) return stbi__err("bad IHDR len","Corrupt PNG"); + s->img_x = stbi__get32be(s); if (s->img_x > (1 << 24)) return stbi__err("too large","Very large image (corrupt?)"); + s->img_y = stbi__get32be(s); if (s->img_y > (1 << 24)) return stbi__err("too large","Very large image (corrupt?)"); + z->depth = stbi__get8(s); if (z->depth != 1 && z->depth != 2 && z->depth != 4 && z->depth != 8 && z->depth != 16) return stbi__err("1/2/4/8/16-bit only","PNG not supported: 1/2/4/8/16-bit only"); + color = stbi__get8(s); if (color > 6) return stbi__err("bad ctype","Corrupt PNG"); + if (color == 3 && z->depth == 16) return stbi__err("bad ctype","Corrupt PNG"); + if (color == 3) pal_img_n = 3; else if (color & 1) return stbi__err("bad ctype","Corrupt PNG"); + comp = stbi__get8(s); if (comp) return stbi__err("bad comp method","Corrupt PNG"); + filter= stbi__get8(s); if (filter) return stbi__err("bad filter method","Corrupt PNG"); + interlace = stbi__get8(s); if (interlace>1) return stbi__err("bad interlace method","Corrupt PNG"); + if (!s->img_x || !s->img_y) return stbi__err("0-pixel image","Corrupt PNG"); + if (!pal_img_n) { + s->img_n = (color & 2 ? 3 : 1) + (color & 4 ? 1 : 0); + if ((1 << 30) / s->img_x / s->img_n < s->img_y) return stbi__err("too large", "Image too large to decode"); + if (scan == STBI__SCAN_header) return 1; + } else { + // if paletted, then pal_n is our final components, and + // img_n is # components to decompress/filter. + s->img_n = 1; + if ((1 << 30) / s->img_x / 4 < s->img_y) return stbi__err("too large","Corrupt PNG"); + // if SCAN_header, have to scan to see if we have a tRNS + } + break; + } + + case STBI__PNG_TYPE('P','L','T','E'): { + if (first) return stbi__err("first not IHDR", "Corrupt PNG"); + if (c.length > 256*3) return stbi__err("invalid PLTE","Corrupt PNG"); + pal_len = c.length / 3; + if (pal_len * 3 != c.length) return stbi__err("invalid PLTE","Corrupt PNG"); + for (i=0; i < pal_len; ++i) { + palette[i*4+0] = stbi__get8(s); + palette[i*4+1] = stbi__get8(s); + palette[i*4+2] = stbi__get8(s); + palette[i*4+3] = 255; + } + break; + } + + case STBI__PNG_TYPE('t','R','N','S'): { + if (first) return stbi__err("first not IHDR", "Corrupt PNG"); + if (z->idata) return stbi__err("tRNS after IDAT","Corrupt PNG"); + if (pal_img_n) { + if (scan == STBI__SCAN_header) { s->img_n = 4; return 1; } + if (pal_len == 0) return stbi__err("tRNS before PLTE","Corrupt PNG"); + if (c.length > pal_len) return stbi__err("bad tRNS len","Corrupt PNG"); + pal_img_n = 4; + for (i=0; i < c.length; ++i) + palette[i*4+3] = stbi__get8(s); + } else { + if (!(s->img_n & 1)) return stbi__err("tRNS with alpha","Corrupt PNG"); + if (c.length != (stbi__uint32) s->img_n*2) return stbi__err("bad tRNS len","Corrupt PNG"); + has_trans = 1; + if (z->depth == 16) { + for (k = 0; k < s->img_n; ++k) tc16[k] = (stbi__uint16)stbi__get16be(s); // copy the values as-is + } else { + for (k = 0; k < s->img_n; ++k) tc[k] = (stbi_uc)(stbi__get16be(s) & 255) * stbi__depth_scale_table[z->depth]; // non 8-bit images will be larger + } + } + break; + } + + case STBI__PNG_TYPE('I','D','A','T'): { + if (first) return stbi__err("first not IHDR", "Corrupt PNG"); + if (pal_img_n && !pal_len) return stbi__err("no PLTE","Corrupt PNG"); + if (scan == STBI__SCAN_header) { s->img_n = pal_img_n; return 1; } + if ((int)(ioff + c.length) < (int)ioff) return 0; + if (ioff + c.length > idata_limit) { + stbi__uint32 idata_limit_old = idata_limit; + stbi_uc *p; + if (idata_limit == 0) idata_limit = c.length > 4096 ? c.length : 4096; + while (ioff + c.length > idata_limit) + idata_limit *= 2; + STBI_NOTUSED(idata_limit_old); + p = (stbi_uc *) STBI_REALLOC_SIZED(z->idata, idata_limit_old, idata_limit); if (p == NULL) return stbi__err("outofmem", "Out of memory"); + z->idata = p; + } + if (!stbi__getn(s, z->idata+ioff,c.length)) return stbi__err("outofdata","Corrupt PNG"); + ioff += c.length; + break; + } + + case STBI__PNG_TYPE('I','E','N','D'): { + stbi__uint32 raw_len, bpl; + if (first) return stbi__err("first not IHDR", "Corrupt PNG"); + if (scan != STBI__SCAN_load) return 1; + if (z->idata == NULL) return stbi__err("no IDAT","Corrupt PNG"); + // initial guess for decoded data size to avoid unnecessary reallocs + bpl = (s->img_x * z->depth + 7) / 8; // bytes per line, per component + raw_len = bpl * s->img_y * s->img_n /* pixels */ + s->img_y /* filter mode per row */; + z->expanded = (stbi_uc *) stbi_zlib_decode_malloc_guesssize_headerflag((char *) z->idata, ioff, raw_len, (int *) &raw_len, !is_iphone); + if (z->expanded == NULL) return 0; // zlib should set error + STBI_FREE(z->idata); z->idata = NULL; + if ((req_comp == s->img_n+1 && req_comp != 3 && !pal_img_n) || has_trans) + s->img_out_n = s->img_n+1; + else + s->img_out_n = s->img_n; + if (!stbi__create_png_image(z, z->expanded, raw_len, s->img_out_n, z->depth, color, interlace)) return 0; + if (has_trans) { + if (z->depth == 16) { + if (!stbi__compute_transparency16(z, tc16, s->img_out_n)) return 0; + } else { + if (!stbi__compute_transparency(z, tc, s->img_out_n)) return 0; + } + } + if (is_iphone && stbi__de_iphone_flag && s->img_out_n > 2) + stbi__de_iphone(z); + if (pal_img_n) { + // pal_img_n == 3 or 4 + s->img_n = pal_img_n; // record the actual colors we had + s->img_out_n = pal_img_n; + if (req_comp >= 3) s->img_out_n = req_comp; + if (!stbi__expand_png_palette(z, palette, pal_len, s->img_out_n)) + return 0; + } else if (has_trans) { + // non-paletted image with tRNS -> source image has (constant) alpha + ++s->img_n; + } + STBI_FREE(z->expanded); z->expanded = NULL; + // end of PNG chunk, read and skip CRC + stbi__get32be(s); + return 1; + } + + default: + // if critical, fail + if (first) return stbi__err("first not IHDR", "Corrupt PNG"); + if ((c.type & (1 << 29)) == 0) { + #ifndef STBI_NO_FAILURE_STRINGS + // not threadsafe + static char invalid_chunk[] = "XXXX PNG chunk not known"; + invalid_chunk[0] = STBI__BYTECAST(c.type >> 24); + invalid_chunk[1] = STBI__BYTECAST(c.type >> 16); + invalid_chunk[2] = STBI__BYTECAST(c.type >> 8); + invalid_chunk[3] = STBI__BYTECAST(c.type >> 0); + #endif + return stbi__err(invalid_chunk, "PNG not supported: unknown PNG chunk type"); + } + stbi__skip(s, c.length); + break; + } + // end of PNG chunk, read and skip CRC + stbi__get32be(s); + } +} + +static void *stbi__do_png(stbi__png *p, int *x, int *y, int *n, int req_comp, stbi__result_info *ri) +{ + void *result=NULL; + if (req_comp < 0 || req_comp > 4) return stbi__errpuc("bad req_comp", "Internal error"); + if (stbi__parse_png_file(p, STBI__SCAN_load, req_comp)) { + if (p->depth < 8) + ri->bits_per_channel = 8; + else + ri->bits_per_channel = p->depth; + result = p->out; + p->out = NULL; + if (req_comp && req_comp != p->s->img_out_n) { + if (ri->bits_per_channel == 8) + result = stbi__convert_format((unsigned char *) result, p->s->img_out_n, req_comp, p->s->img_x, p->s->img_y); + else + result = stbi__convert_format16((stbi__uint16 *) result, p->s->img_out_n, req_comp, p->s->img_x, p->s->img_y); + p->s->img_out_n = req_comp; + if (result == NULL) return result; + } + *x = p->s->img_x; + *y = p->s->img_y; + if (n) *n = p->s->img_n; + } + STBI_FREE(p->out); p->out = NULL; + STBI_FREE(p->expanded); p->expanded = NULL; + STBI_FREE(p->idata); p->idata = NULL; + + return result; +} + +static void *stbi__png_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri) +{ + stbi__png p; + p.s = s; + return stbi__do_png(&p, x,y,comp,req_comp, ri); +} + +static int stbi__png_test(stbi__context *s) +{ + int r; + r = stbi__check_png_header(s); + stbi__rewind(s); + return r; +} + +static int stbi__png_info_raw(stbi__png *p, int *x, int *y, int *comp) +{ + if (!stbi__parse_png_file(p, STBI__SCAN_header, 0)) { + stbi__rewind( p->s ); + return 0; + } + if (x) *x = p->s->img_x; + if (y) *y = p->s->img_y; + if (comp) *comp = p->s->img_n; + return 1; +} + +static int stbi__png_info(stbi__context *s, int *x, int *y, int *comp) +{ + stbi__png p; + p.s = s; + return stbi__png_info_raw(&p, x, y, comp); +} + +static int stbi__png_is16(stbi__context *s) +{ + stbi__png p; + p.s = s; + if (!stbi__png_info_raw(&p, NULL, NULL, NULL)) + return 0; + if (p.depth != 16) { + stbi__rewind(p.s); + return 0; + } + return 1; +} +#endif + +// Microsoft/Windows BMP image + +#ifndef STBI_NO_BMP +static int stbi__bmp_test_raw(stbi__context *s) +{ + int r; + int sz; + if (stbi__get8(s) != 'B') return 0; + if (stbi__get8(s) != 'M') return 0; + stbi__get32le(s); // discard filesize + stbi__get16le(s); // discard reserved + stbi__get16le(s); // discard reserved + stbi__get32le(s); // discard data offset + sz = stbi__get32le(s); + r = (sz == 12 || sz == 40 || sz == 56 || sz == 108 || sz == 124); + return r; +} + +static int stbi__bmp_test(stbi__context *s) +{ + int r = stbi__bmp_test_raw(s); + stbi__rewind(s); + return r; +} + + +// returns 0..31 for the highest set bit +static int stbi__high_bit(unsigned int z) +{ + int n=0; + if (z == 0) return -1; + if (z >= 0x10000) { n += 16; z >>= 16; } + if (z >= 0x00100) { n += 8; z >>= 8; } + if (z >= 0x00010) { n += 4; z >>= 4; } + if (z >= 0x00004) { n += 2; z >>= 2; } + if (z >= 0x00002) { n += 1;/* >>= 1;*/ } + return n; +} + +static int stbi__bitcount(unsigned int a) +{ + a = (a & 0x55555555) + ((a >> 1) & 0x55555555); // max 2 + a = (a & 0x33333333) + ((a >> 2) & 0x33333333); // max 4 + a = (a + (a >> 4)) & 0x0f0f0f0f; // max 8 per 4, now 8 bits + a = (a + (a >> 8)); // max 16 per 8 bits + a = (a + (a >> 16)); // max 32 per 8 bits + return a & 0xff; +} + +// extract an arbitrarily-aligned N-bit value (N=bits) +// from v, and then make it 8-bits long and fractionally +// extend it to full full range. +static int stbi__shiftsigned(unsigned int v, int shift, int bits) +{ + static unsigned int mul_table[9] = { + 0, + 0xff/*0b11111111*/, 0x55/*0b01010101*/, 0x49/*0b01001001*/, 0x11/*0b00010001*/, + 0x21/*0b00100001*/, 0x41/*0b01000001*/, 0x81/*0b10000001*/, 0x01/*0b00000001*/, + }; + static unsigned int shift_table[9] = { + 0, 0,0,1,0,2,4,6,0, + }; + if (shift < 0) + v <<= -shift; + else + v >>= shift; + STBI_ASSERT(v < 256); + v >>= (8-bits); + STBI_ASSERT(bits >= 0 && bits <= 8); + return (int) ((unsigned) v * mul_table[bits]) >> shift_table[bits]; +} + +typedef struct +{ + int bpp, offset, hsz; + unsigned int mr,mg,mb,ma, all_a; + int extra_read; +} stbi__bmp_data; + +static void *stbi__bmp_parse_header(stbi__context *s, stbi__bmp_data *info) +{ + int hsz; + if (stbi__get8(s) != 'B' || stbi__get8(s) != 'M') return stbi__errpuc("not BMP", "Corrupt BMP"); + stbi__get32le(s); // discard filesize + stbi__get16le(s); // discard reserved + stbi__get16le(s); // discard reserved + info->offset = stbi__get32le(s); + info->hsz = hsz = stbi__get32le(s); + info->mr = info->mg = info->mb = info->ma = 0; + info->extra_read = 14; + + if (hsz != 12 && hsz != 40 && hsz != 56 && hsz != 108 && hsz != 124) return stbi__errpuc("unknown BMP", "BMP type not supported: unknown"); + if (hsz == 12) { + s->img_x = stbi__get16le(s); + s->img_y = stbi__get16le(s); + } else { + s->img_x = stbi__get32le(s); + s->img_y = stbi__get32le(s); + } + if (stbi__get16le(s) != 1) return stbi__errpuc("bad BMP", "bad BMP"); + info->bpp = stbi__get16le(s); + if (hsz != 12) { + int compress = stbi__get32le(s); + if (compress == 1 || compress == 2) return stbi__errpuc("BMP RLE", "BMP type not supported: RLE"); + stbi__get32le(s); // discard sizeof + stbi__get32le(s); // discard hres + stbi__get32le(s); // discard vres + stbi__get32le(s); // discard colorsused + stbi__get32le(s); // discard max important + if (hsz == 40 || hsz == 56) { + if (hsz == 56) { + stbi__get32le(s); + stbi__get32le(s); + stbi__get32le(s); + stbi__get32le(s); + } + if (info->bpp == 16 || info->bpp == 32) { + if (compress == 0) { + if (info->bpp == 32) { + info->mr = 0xffu << 16; + info->mg = 0xffu << 8; + info->mb = 0xffu << 0; + info->ma = 0xffu << 24; + info->all_a = 0; // if all_a is 0 at end, then we loaded alpha channel but it was all 0 + } else { + info->mr = 31u << 10; + info->mg = 31u << 5; + info->mb = 31u << 0; + } + } else if (compress == 3) { + info->mr = stbi__get32le(s); + info->mg = stbi__get32le(s); + info->mb = stbi__get32le(s); + info->extra_read += 12; + // not documented, but generated by photoshop and handled by mspaint + if (info->mr == info->mg && info->mg == info->mb) { + // ?!?!? + return stbi__errpuc("bad BMP", "bad BMP"); + } + } else + return stbi__errpuc("bad BMP", "bad BMP"); + } + } else { + int i; + if (hsz != 108 && hsz != 124) + return stbi__errpuc("bad BMP", "bad BMP"); + info->mr = stbi__get32le(s); + info->mg = stbi__get32le(s); + info->mb = stbi__get32le(s); + info->ma = stbi__get32le(s); + stbi__get32le(s); // discard color space + for (i=0; i < 12; ++i) + stbi__get32le(s); // discard color space parameters + if (hsz == 124) { + stbi__get32le(s); // discard rendering intent + stbi__get32le(s); // discard offset of profile data + stbi__get32le(s); // discard size of profile data + stbi__get32le(s); // discard reserved + } + } + } + return (void *) 1; +} + + +static void *stbi__bmp_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri) +{ + stbi_uc *out; + unsigned int mr=0,mg=0,mb=0,ma=0, all_a; + stbi_uc pal[256][4]; + int psize=0,i,j,width; + int flip_vertically, pad, target; + stbi__bmp_data info; + STBI_NOTUSED(ri); + + info.all_a = 255; + if (stbi__bmp_parse_header(s, &info) == NULL) + return NULL; // error code already set + + flip_vertically = ((int) s->img_y) > 0; + s->img_y = abs((int) s->img_y); + + mr = info.mr; + mg = info.mg; + mb = info.mb; + ma = info.ma; + all_a = info.all_a; + + if (info.hsz == 12) { + if (info.bpp < 24) + psize = (info.offset - info.extra_read - 24) / 3; + } else { + if (info.bpp < 16) + psize = (info.offset - info.extra_read - info.hsz) >> 2; + } + if (psize == 0) { + STBI_ASSERT(info.offset == (s->img_buffer - s->buffer_start)); + } + + if (info.bpp == 24 && ma == 0xff000000) + s->img_n = 3; + else + s->img_n = ma ? 4 : 3; + if (req_comp && req_comp >= 3) // we can directly decode 3 or 4 + target = req_comp; + else + target = s->img_n; // if they want monochrome, we'll post-convert + + // sanity-check size + if (!stbi__mad3sizes_valid(target, s->img_x, s->img_y, 0)) + return stbi__errpuc("too large", "Corrupt BMP"); + + out = (stbi_uc *) stbi__malloc_mad3(target, s->img_x, s->img_y, 0); + if (!out) return stbi__errpuc("outofmem", "Out of memory"); + if (info.bpp < 16) { + int z=0; + if (psize == 0 || psize > 256) { STBI_FREE(out); return stbi__errpuc("invalid", "Corrupt BMP"); } + for (i=0; i < psize; ++i) { + pal[i][2] = stbi__get8(s); + pal[i][1] = stbi__get8(s); + pal[i][0] = stbi__get8(s); + if (info.hsz != 12) stbi__get8(s); + pal[i][3] = 255; + } + stbi__skip(s, info.offset - info.extra_read - info.hsz - psize * (info.hsz == 12 ? 3 : 4)); + if (info.bpp == 1) width = (s->img_x + 7) >> 3; + else if (info.bpp == 4) width = (s->img_x + 1) >> 1; + else if (info.bpp == 8) width = s->img_x; + else { STBI_FREE(out); return stbi__errpuc("bad bpp", "Corrupt BMP"); } + pad = (-width)&3; + if (info.bpp == 1) { + for (j=0; j < (int) s->img_y; ++j) { + int bit_offset = 7, v = stbi__get8(s); + for (i=0; i < (int) s->img_x; ++i) { + int color = (v>>bit_offset)&0x1; + out[z++] = pal[color][0]; + out[z++] = pal[color][1]; + out[z++] = pal[color][2]; + if (target == 4) out[z++] = 255; + if (i+1 == (int) s->img_x) break; + if((--bit_offset) < 0) { + bit_offset = 7; + v = stbi__get8(s); + } + } + stbi__skip(s, pad); + } + } else { + for (j=0; j < (int) s->img_y; ++j) { + for (i=0; i < (int) s->img_x; i += 2) { + int v=stbi__get8(s),v2=0; + if (info.bpp == 4) { + v2 = v & 15; + v >>= 4; + } + out[z++] = pal[v][0]; + out[z++] = pal[v][1]; + out[z++] = pal[v][2]; + if (target == 4) out[z++] = 255; + if (i+1 == (int) s->img_x) break; + v = (info.bpp == 8) ? stbi__get8(s) : v2; + out[z++] = pal[v][0]; + out[z++] = pal[v][1]; + out[z++] = pal[v][2]; + if (target == 4) out[z++] = 255; + } + stbi__skip(s, pad); + } + } + } else { + int rshift=0,gshift=0,bshift=0,ashift=0,rcount=0,gcount=0,bcount=0,acount=0; + int z = 0; + int easy=0; + stbi__skip(s, info.offset - info.extra_read - info.hsz); + if (info.bpp == 24) width = 3 * s->img_x; + else if (info.bpp == 16) width = 2*s->img_x; + else /* bpp = 32 and pad = 0 */ width=0; + pad = (-width) & 3; + if (info.bpp == 24) { + easy = 1; + } else if (info.bpp == 32) { + if (mb == 0xff && mg == 0xff00 && mr == 0x00ff0000 && ma == 0xff000000) + easy = 2; + } + if (!easy) { + if (!mr || !mg || !mb) { STBI_FREE(out); return stbi__errpuc("bad masks", "Corrupt BMP"); } + // right shift amt to put high bit in position #7 + rshift = stbi__high_bit(mr)-7; rcount = stbi__bitcount(mr); + gshift = stbi__high_bit(mg)-7; gcount = stbi__bitcount(mg); + bshift = stbi__high_bit(mb)-7; bcount = stbi__bitcount(mb); + ashift = stbi__high_bit(ma)-7; acount = stbi__bitcount(ma); + } + for (j=0; j < (int) s->img_y; ++j) { + if (easy) { + for (i=0; i < (int) s->img_x; ++i) { + unsigned char a; + out[z+2] = stbi__get8(s); + out[z+1] = stbi__get8(s); + out[z+0] = stbi__get8(s); + z += 3; + a = (easy == 2 ? stbi__get8(s) : 255); + all_a |= a; + if (target == 4) out[z++] = a; + } + } else { + int bpp = info.bpp; + for (i=0; i < (int) s->img_x; ++i) { + stbi__uint32 v = (bpp == 16 ? (stbi__uint32) stbi__get16le(s) : stbi__get32le(s)); + unsigned int a; + out[z++] = STBI__BYTECAST(stbi__shiftsigned(v & mr, rshift, rcount)); + out[z++] = STBI__BYTECAST(stbi__shiftsigned(v & mg, gshift, gcount)); + out[z++] = STBI__BYTECAST(stbi__shiftsigned(v & mb, bshift, bcount)); + a = (ma ? stbi__shiftsigned(v & ma, ashift, acount) : 255); + all_a |= a; + if (target == 4) out[z++] = STBI__BYTECAST(a); + } + } + stbi__skip(s, pad); + } + } + + // if alpha channel is all 0s, replace with all 255s + if (target == 4 && all_a == 0) + for (i=4*s->img_x*s->img_y-1; i >= 0; i -= 4) + out[i] = 255; + + if (flip_vertically) { + stbi_uc t; + for (j=0; j < (int) s->img_y>>1; ++j) { + stbi_uc *p1 = out + j *s->img_x*target; + stbi_uc *p2 = out + (s->img_y-1-j)*s->img_x*target; + for (i=0; i < (int) s->img_x*target; ++i) { + t = p1[i]; p1[i] = p2[i]; p2[i] = t; + } + } + } + + if (req_comp && req_comp != target) { + out = stbi__convert_format(out, target, req_comp, s->img_x, s->img_y); + if (out == NULL) return out; // stbi__convert_format frees input on failure + } + + *x = s->img_x; + *y = s->img_y; + if (comp) *comp = s->img_n; + return out; +} +#endif + +// Targa Truevision - TGA +// by Jonathan Dummer +#ifndef STBI_NO_TGA +// returns STBI_rgb or whatever, 0 on error +static int stbi__tga_get_comp(int bits_per_pixel, int is_grey, int* is_rgb16) +{ + // only RGB or RGBA (incl. 16bit) or grey allowed + if (is_rgb16) *is_rgb16 = 0; + switch(bits_per_pixel) { + case 8: return STBI_grey; + case 16: if(is_grey) return STBI_grey_alpha; + // fallthrough + case 15: if(is_rgb16) *is_rgb16 = 1; + return STBI_rgb; + case 24: // fallthrough + case 32: return bits_per_pixel/8; + default: return 0; + } +} + +static int stbi__tga_info(stbi__context *s, int *x, int *y, int *comp) +{ + int tga_w, tga_h, tga_comp, tga_image_type, tga_bits_per_pixel, tga_colormap_bpp; + int sz, tga_colormap_type; + stbi__get8(s); // discard Offset + tga_colormap_type = stbi__get8(s); // colormap type + if( tga_colormap_type > 1 ) { + stbi__rewind(s); + return 0; // only RGB or indexed allowed + } + tga_image_type = stbi__get8(s); // image type + if ( tga_colormap_type == 1 ) { // colormapped (paletted) image + if (tga_image_type != 1 && tga_image_type != 9) { + stbi__rewind(s); + return 0; + } + stbi__skip(s,4); // skip index of first colormap entry and number of entries + sz = stbi__get8(s); // check bits per palette color entry + if ( (sz != 8) && (sz != 15) && (sz != 16) && (sz != 24) && (sz != 32) ) { + stbi__rewind(s); + return 0; + } + stbi__skip(s,4); // skip image x and y origin + tga_colormap_bpp = sz; + } else { // "normal" image w/o colormap - only RGB or grey allowed, +/- RLE + if ( (tga_image_type != 2) && (tga_image_type != 3) && (tga_image_type != 10) && (tga_image_type != 11) ) { + stbi__rewind(s); + return 0; // only RGB or grey allowed, +/- RLE + } + stbi__skip(s,9); // skip colormap specification and image x/y origin + tga_colormap_bpp = 0; + } + tga_w = stbi__get16le(s); + if( tga_w < 1 ) { + stbi__rewind(s); + return 0; // test width + } + tga_h = stbi__get16le(s); + if( tga_h < 1 ) { + stbi__rewind(s); + return 0; // test height + } + tga_bits_per_pixel = stbi__get8(s); // bits per pixel + stbi__get8(s); // ignore alpha bits + if (tga_colormap_bpp != 0) { + if((tga_bits_per_pixel != 8) && (tga_bits_per_pixel != 16)) { + // when using a colormap, tga_bits_per_pixel is the size of the indexes + // I don't think anything but 8 or 16bit indexes makes sense + stbi__rewind(s); + return 0; + } + tga_comp = stbi__tga_get_comp(tga_colormap_bpp, 0, NULL); + } else { + tga_comp = stbi__tga_get_comp(tga_bits_per_pixel, (tga_image_type == 3) || (tga_image_type == 11), NULL); + } + if(!tga_comp) { + stbi__rewind(s); + return 0; + } + if (x) *x = tga_w; + if (y) *y = tga_h; + if (comp) *comp = tga_comp; + return 1; // seems to have passed everything +} + +static int stbi__tga_test(stbi__context *s) +{ + int res = 0; + int sz, tga_color_type; + stbi__get8(s); // discard Offset + tga_color_type = stbi__get8(s); // color type + if ( tga_color_type > 1 ) goto errorEnd; // only RGB or indexed allowed + sz = stbi__get8(s); // image type + if ( tga_color_type == 1 ) { // colormapped (paletted) image + if (sz != 1 && sz != 9) goto errorEnd; // colortype 1 demands image type 1 or 9 + stbi__skip(s,4); // skip index of first colormap entry and number of entries + sz = stbi__get8(s); // check bits per palette color entry + if ( (sz != 8) && (sz != 15) && (sz != 16) && (sz != 24) && (sz != 32) ) goto errorEnd; + stbi__skip(s,4); // skip image x and y origin + } else { // "normal" image w/o colormap + if ( (sz != 2) && (sz != 3) && (sz != 10) && (sz != 11) ) goto errorEnd; // only RGB or grey allowed, +/- RLE + stbi__skip(s,9); // skip colormap specification and image x/y origin + } + if ( stbi__get16le(s) < 1 ) goto errorEnd; // test width + if ( stbi__get16le(s) < 1 ) goto errorEnd; // test height + sz = stbi__get8(s); // bits per pixel + if ( (tga_color_type == 1) && (sz != 8) && (sz != 16) ) goto errorEnd; // for colormapped images, bpp is size of an index + if ( (sz != 8) && (sz != 15) && (sz != 16) && (sz != 24) && (sz != 32) ) goto errorEnd; + + res = 1; // if we got this far, everything's good and we can return 1 instead of 0 + +errorEnd: + stbi__rewind(s); + return res; +} + +// read 16bit value and convert to 24bit RGB +static void stbi__tga_read_rgb16(stbi__context *s, stbi_uc* out) +{ + stbi__uint16 px = (stbi__uint16)stbi__get16le(s); + stbi__uint16 fiveBitMask = 31; + // we have 3 channels with 5bits each + int r = (px >> 10) & fiveBitMask; + int g = (px >> 5) & fiveBitMask; + int b = px & fiveBitMask; + // Note that this saves the data in RGB(A) order, so it doesn't need to be swapped later + out[0] = (stbi_uc)((r * 255)/31); + out[1] = (stbi_uc)((g * 255)/31); + out[2] = (stbi_uc)((b * 255)/31); + + // some people claim that the most significant bit might be used for alpha + // (possibly if an alpha-bit is set in the "image descriptor byte") + // but that only made 16bit test images completely translucent.. + // so let's treat all 15 and 16bit TGAs as RGB with no alpha. +} + +static void *stbi__tga_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri) +{ + // read in the TGA header stuff + int tga_offset = stbi__get8(s); + int tga_indexed = stbi__get8(s); + int tga_image_type = stbi__get8(s); + int tga_is_RLE = 0; + int tga_palette_start = stbi__get16le(s); + int tga_palette_len = stbi__get16le(s); + int tga_palette_bits = stbi__get8(s); + int tga_x_origin = stbi__get16le(s); + int tga_y_origin = stbi__get16le(s); + int tga_width = stbi__get16le(s); + int tga_height = stbi__get16le(s); + int tga_bits_per_pixel = stbi__get8(s); + int tga_comp, tga_rgb16=0; + int tga_inverted = stbi__get8(s); + // int tga_alpha_bits = tga_inverted & 15; // the 4 lowest bits - unused (useless?) + // image data + unsigned char *tga_data; + unsigned char *tga_palette = NULL; + int i, j; + unsigned char raw_data[4] = {0}; + int RLE_count = 0; + int RLE_repeating = 0; + int read_next_pixel = 1; + STBI_NOTUSED(ri); + STBI_NOTUSED(tga_x_origin); // @TODO + STBI_NOTUSED(tga_y_origin); // @TODO + + // do a tiny bit of precessing + if ( tga_image_type >= 8 ) + { + tga_image_type -= 8; + tga_is_RLE = 1; + } + tga_inverted = 1 - ((tga_inverted >> 5) & 1); + + // If I'm paletted, then I'll use the number of bits from the palette + if ( tga_indexed ) tga_comp = stbi__tga_get_comp(tga_palette_bits, 0, &tga_rgb16); + else tga_comp = stbi__tga_get_comp(tga_bits_per_pixel, (tga_image_type == 3), &tga_rgb16); + + if(!tga_comp) // shouldn't really happen, stbi__tga_test() should have ensured basic consistency + return stbi__errpuc("bad format", "Can't find out TGA pixelformat"); + + // tga info + *x = tga_width; + *y = tga_height; + if (comp) *comp = tga_comp; + + if (!stbi__mad3sizes_valid(tga_width, tga_height, tga_comp, 0)) + return stbi__errpuc("too large", "Corrupt TGA"); + + tga_data = (unsigned char*)stbi__malloc_mad3(tga_width, tga_height, tga_comp, 0); + if (!tga_data) return stbi__errpuc("outofmem", "Out of memory"); + + // skip to the data's starting position (offset usually = 0) + stbi__skip(s, tga_offset ); + + if ( !tga_indexed && !tga_is_RLE && !tga_rgb16 ) { + for (i=0; i < tga_height; ++i) { + int row = tga_inverted ? tga_height -i - 1 : i; + stbi_uc *tga_row = tga_data + row*tga_width*tga_comp; + stbi__getn(s, tga_row, tga_width * tga_comp); + } + } else { + // do I need to load a palette? + if ( tga_indexed) + { + // any data to skip? (offset usually = 0) + stbi__skip(s, tga_palette_start ); + // load the palette + tga_palette = (unsigned char*)stbi__malloc_mad2(tga_palette_len, tga_comp, 0); + if (!tga_palette) { + STBI_FREE(tga_data); + return stbi__errpuc("outofmem", "Out of memory"); + } + if (tga_rgb16) { + stbi_uc *pal_entry = tga_palette; + STBI_ASSERT(tga_comp == STBI_rgb); + for (i=0; i < tga_palette_len; ++i) { + stbi__tga_read_rgb16(s, pal_entry); + pal_entry += tga_comp; + } + } else if (!stbi__getn(s, tga_palette, tga_palette_len * tga_comp)) { + STBI_FREE(tga_data); + STBI_FREE(tga_palette); + return stbi__errpuc("bad palette", "Corrupt TGA"); + } + } + // load the data + for (i=0; i < tga_width * tga_height; ++i) + { + // if I'm in RLE mode, do I need to get a RLE stbi__pngchunk? + if ( tga_is_RLE ) + { + if ( RLE_count == 0 ) + { + // yep, get the next byte as a RLE command + int RLE_cmd = stbi__get8(s); + RLE_count = 1 + (RLE_cmd & 127); + RLE_repeating = RLE_cmd >> 7; + read_next_pixel = 1; + } else if ( !RLE_repeating ) + { + read_next_pixel = 1; + } + } else + { + read_next_pixel = 1; + } + // OK, if I need to read a pixel, do it now + if ( read_next_pixel ) + { + // load however much data we did have + if ( tga_indexed ) + { + // read in index, then perform the lookup + int pal_idx = (tga_bits_per_pixel == 8) ? stbi__get8(s) : stbi__get16le(s); + if ( pal_idx >= tga_palette_len ) { + // invalid index + pal_idx = 0; + } + pal_idx *= tga_comp; + for (j = 0; j < tga_comp; ++j) { + raw_data[j] = tga_palette[pal_idx+j]; + } + } else if(tga_rgb16) { + STBI_ASSERT(tga_comp == STBI_rgb); + stbi__tga_read_rgb16(s, raw_data); + } else { + // read in the data raw + for (j = 0; j < tga_comp; ++j) { + raw_data[j] = stbi__get8(s); + } + } + // clear the reading flag for the next pixel + read_next_pixel = 0; + } // end of reading a pixel + + // copy data + for (j = 0; j < tga_comp; ++j) + tga_data[i*tga_comp+j] = raw_data[j]; + + // in case we're in RLE mode, keep counting down + --RLE_count; + } + // do I need to invert the image? + if ( tga_inverted ) + { + for (j = 0; j*2 < tga_height; ++j) + { + int index1 = j * tga_width * tga_comp; + int index2 = (tga_height - 1 - j) * tga_width * tga_comp; + for (i = tga_width * tga_comp; i > 0; --i) + { + unsigned char temp = tga_data[index1]; + tga_data[index1] = tga_data[index2]; + tga_data[index2] = temp; + ++index1; + ++index2; + } + } + } + // clear my palette, if I had one + if ( tga_palette != NULL ) + { + STBI_FREE( tga_palette ); + } + } + + // swap RGB - if the source data was RGB16, it already is in the right order + if (tga_comp >= 3 && !tga_rgb16) + { + unsigned char* tga_pixel = tga_data; + for (i=0; i < tga_width * tga_height; ++i) + { + unsigned char temp = tga_pixel[0]; + tga_pixel[0] = tga_pixel[2]; + tga_pixel[2] = temp; + tga_pixel += tga_comp; + } + } + + // convert to target component count + if (req_comp && req_comp != tga_comp) + tga_data = stbi__convert_format(tga_data, tga_comp, req_comp, tga_width, tga_height); + + // the things I do to get rid of an error message, and yet keep + // Microsoft's C compilers happy... [8^( + tga_palette_start = tga_palette_len = tga_palette_bits = + tga_x_origin = tga_y_origin = 0; + STBI_NOTUSED(tga_palette_start); + // OK, done + return tga_data; +} +#endif + +// ************************************************************************************************* +// Photoshop PSD loader -- PD by Thatcher Ulrich, integration by Nicolas Schulz, tweaked by STB + +#ifndef STBI_NO_PSD +static int stbi__psd_test(stbi__context *s) +{ + int r = (stbi__get32be(s) == 0x38425053); + stbi__rewind(s); + return r; +} + +static int stbi__psd_decode_rle(stbi__context *s, stbi_uc *p, int pixelCount) +{ + int count, nleft, len; + + count = 0; + while ((nleft = pixelCount - count) > 0) { + len = stbi__get8(s); + if (len == 128) { + // No-op. + } else if (len < 128) { + // Copy next len+1 bytes literally. + len++; + if (len > nleft) return 0; // corrupt data + count += len; + while (len) { + *p = stbi__get8(s); + p += 4; + len--; + } + } else if (len > 128) { + stbi_uc val; + // Next -len+1 bytes in the dest are replicated from next source byte. + // (Interpret len as a negative 8-bit int.) + len = 257 - len; + if (len > nleft) return 0; // corrupt data + val = stbi__get8(s); + count += len; + while (len) { + *p = val; + p += 4; + len--; + } + } + } + + return 1; +} + +static void *stbi__psd_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri, int bpc) +{ + int pixelCount; + int channelCount, compression; + int channel, i; + int bitdepth; + int w,h; + stbi_uc *out; + STBI_NOTUSED(ri); + + // Check identifier + if (stbi__get32be(s) != 0x38425053) // "8BPS" + return stbi__errpuc("not PSD", "Corrupt PSD image"); + + // Check file type version. + if (stbi__get16be(s) != 1) + return stbi__errpuc("wrong version", "Unsupported version of PSD image"); + + // Skip 6 reserved bytes. + stbi__skip(s, 6 ); + + // Read the number of channels (R, G, B, A, etc). + channelCount = stbi__get16be(s); + if (channelCount < 0 || channelCount > 16) + return stbi__errpuc("wrong channel count", "Unsupported number of channels in PSD image"); + + // Read the rows and columns of the image. + h = stbi__get32be(s); + w = stbi__get32be(s); + + // Make sure the depth is 8 bits. + bitdepth = stbi__get16be(s); + if (bitdepth != 8 && bitdepth != 16) + return stbi__errpuc("unsupported bit depth", "PSD bit depth is not 8 or 16 bit"); + + // Make sure the color mode is RGB. + // Valid options are: + // 0: Bitmap + // 1: Grayscale + // 2: Indexed color + // 3: RGB color + // 4: CMYK color + // 7: Multichannel + // 8: Duotone + // 9: Lab color + if (stbi__get16be(s) != 3) + return stbi__errpuc("wrong color format", "PSD is not in RGB color format"); + + // Skip the Mode Data. (It's the palette for indexed color; other info for other modes.) + stbi__skip(s,stbi__get32be(s) ); + + // Skip the image resources. (resolution, pen tool paths, etc) + stbi__skip(s, stbi__get32be(s) ); + + // Skip the reserved data. + stbi__skip(s, stbi__get32be(s) ); + + // Find out if the data is compressed. + // Known values: + // 0: no compression + // 1: RLE compressed + compression = stbi__get16be(s); + if (compression > 1) + return stbi__errpuc("bad compression", "PSD has an unknown compression format"); + + // Check size + if (!stbi__mad3sizes_valid(4, w, h, 0)) + return stbi__errpuc("too large", "Corrupt PSD"); + + // Create the destination image. + + if (!compression && bitdepth == 16 && bpc == 16) { + out = (stbi_uc *) stbi__malloc_mad3(8, w, h, 0); + ri->bits_per_channel = 16; + } else + out = (stbi_uc *) stbi__malloc(4 * w*h); + + if (!out) return stbi__errpuc("outofmem", "Out of memory"); + pixelCount = w*h; + + // Initialize the data to zero. + //memset( out, 0, pixelCount * 4 ); + + // Finally, the image data. + if (compression) { + // RLE as used by .PSD and .TIFF + // Loop until you get the number of unpacked bytes you are expecting: + // Read the next source byte into n. + // If n is between 0 and 127 inclusive, copy the next n+1 bytes literally. + // Else if n is between -127 and -1 inclusive, copy the next byte -n+1 times. + // Else if n is 128, noop. + // Endloop + + // The RLE-compressed data is preceded by a 2-byte data count for each row in the data, + // which we're going to just skip. + stbi__skip(s, h * channelCount * 2 ); + + // Read the RLE data by channel. + for (channel = 0; channel < 4; channel++) { + stbi_uc *p; + + p = out+channel; + if (channel >= channelCount) { + // Fill this channel with default data. + for (i = 0; i < pixelCount; i++, p += 4) + *p = (channel == 3 ? 255 : 0); + } else { + // Read the RLE data. + if (!stbi__psd_decode_rle(s, p, pixelCount)) { + STBI_FREE(out); + return stbi__errpuc("corrupt", "bad RLE data"); + } + } + } + + } else { + // We're at the raw image data. It's each channel in order (Red, Green, Blue, Alpha, ...) + // where each channel consists of an 8-bit (or 16-bit) value for each pixel in the image. + + // Read the data by channel. + for (channel = 0; channel < 4; channel++) { + if (channel >= channelCount) { + // Fill this channel with default data. + if (bitdepth == 16 && bpc == 16) { + stbi__uint16 *q = ((stbi__uint16 *) out) + channel; + stbi__uint16 val = channel == 3 ? 65535 : 0; + for (i = 0; i < pixelCount; i++, q += 4) + *q = val; + } else { + stbi_uc *p = out+channel; + stbi_uc val = channel == 3 ? 255 : 0; + for (i = 0; i < pixelCount; i++, p += 4) + *p = val; + } + } else { + if (ri->bits_per_channel == 16) { // output bpc + stbi__uint16 *q = ((stbi__uint16 *) out) + channel; + for (i = 0; i < pixelCount; i++, q += 4) + *q = (stbi__uint16) stbi__get16be(s); + } else { + stbi_uc *p = out+channel; + if (bitdepth == 16) { // input bpc + for (i = 0; i < pixelCount; i++, p += 4) + *p = (stbi_uc) (stbi__get16be(s) >> 8); + } else { + for (i = 0; i < pixelCount; i++, p += 4) + *p = stbi__get8(s); + } + } + } + } + } + + // remove weird white matte from PSD + if (channelCount >= 4) { + if (ri->bits_per_channel == 16) { + for (i=0; i < w*h; ++i) { + stbi__uint16 *pixel = (stbi__uint16 *) out + 4*i; + if (pixel[3] != 0 && pixel[3] != 65535) { + float a = pixel[3] / 65535.0f; + float ra = 1.0f / a; + float inv_a = 65535.0f * (1 - ra); + pixel[0] = (stbi__uint16) (pixel[0]*ra + inv_a); + pixel[1] = (stbi__uint16) (pixel[1]*ra + inv_a); + pixel[2] = (stbi__uint16) (pixel[2]*ra + inv_a); + } + } + } else { + for (i=0; i < w*h; ++i) { + unsigned char *pixel = out + 4*i; + if (pixel[3] != 0 && pixel[3] != 255) { + float a = pixel[3] / 255.0f; + float ra = 1.0f / a; + float inv_a = 255.0f * (1 - ra); + pixel[0] = (unsigned char) (pixel[0]*ra + inv_a); + pixel[1] = (unsigned char) (pixel[1]*ra + inv_a); + pixel[2] = (unsigned char) (pixel[2]*ra + inv_a); + } + } + } + } + + // convert to desired output format + if (req_comp && req_comp != 4) { + if (ri->bits_per_channel == 16) + out = (stbi_uc *) stbi__convert_format16((stbi__uint16 *) out, 4, req_comp, w, h); + else + out = stbi__convert_format(out, 4, req_comp, w, h); + if (out == NULL) return out; // stbi__convert_format frees input on failure + } + + if (comp) *comp = 4; + *y = h; + *x = w; + + return out; +} +#endif + +// ************************************************************************************************* +// Softimage PIC loader +// by Tom Seddon +// +// See http://softimage.wiki.softimage.com/index.php/INFO:_PIC_file_format +// See http://ozviz.wasp.uwa.edu.au/~pbourke/dataformats/softimagepic/ + +#ifndef STBI_NO_PIC +static int stbi__pic_is4(stbi__context *s,const char *str) +{ + int i; + for (i=0; i<4; ++i) + if (stbi__get8(s) != (stbi_uc)str[i]) + return 0; + + return 1; +} + +static int stbi__pic_test_core(stbi__context *s) +{ + int i; + + if (!stbi__pic_is4(s,"\x53\x80\xF6\x34")) + return 0; + + for(i=0;i<84;++i) + stbi__get8(s); + + if (!stbi__pic_is4(s,"PICT")) + return 0; + + return 1; +} + +typedef struct +{ + stbi_uc size,type,channel; +} stbi__pic_packet; + +static stbi_uc *stbi__readval(stbi__context *s, int channel, stbi_uc *dest) +{ + int mask=0x80, i; + + for (i=0; i<4; ++i, mask>>=1) { + if (channel & mask) { + if (stbi__at_eof(s)) return stbi__errpuc("bad file","PIC file too short"); + dest[i]=stbi__get8(s); + } + } + + return dest; +} + +static void stbi__copyval(int channel,stbi_uc *dest,const stbi_uc *src) +{ + int mask=0x80,i; + + for (i=0;i<4; ++i, mask>>=1) + if (channel&mask) + dest[i]=src[i]; +} + +static stbi_uc *stbi__pic_load_core(stbi__context *s,int width,int height,int *comp, stbi_uc *result) +{ + int act_comp=0,num_packets=0,y,chained; + stbi__pic_packet packets[10]; + + // this will (should...) cater for even some bizarre stuff like having data + // for the same channel in multiple packets. + do { + stbi__pic_packet *packet; + + if (num_packets==sizeof(packets)/sizeof(packets[0])) + return stbi__errpuc("bad format","too many packets"); + + packet = &packets[num_packets++]; + + chained = stbi__get8(s); + packet->size = stbi__get8(s); + packet->type = stbi__get8(s); + packet->channel = stbi__get8(s); + + act_comp |= packet->channel; + + if (stbi__at_eof(s)) return stbi__errpuc("bad file","file too short (reading packets)"); + if (packet->size != 8) return stbi__errpuc("bad format","packet isn't 8bpp"); + } while (chained); + + *comp = (act_comp & 0x10 ? 4 : 3); // has alpha channel? + + for(y=0; ytype) { + default: + return stbi__errpuc("bad format","packet has bad compression type"); + + case 0: {//uncompressed + int x; + + for(x=0;xchannel,dest)) + return 0; + break; + } + + case 1://Pure RLE + { + int left=width, i; + + while (left>0) { + stbi_uc count,value[4]; + + count=stbi__get8(s); + if (stbi__at_eof(s)) return stbi__errpuc("bad file","file too short (pure read count)"); + + if (count > left) + count = (stbi_uc) left; + + if (!stbi__readval(s,packet->channel,value)) return 0; + + for(i=0; ichannel,dest,value); + left -= count; + } + } + break; + + case 2: {//Mixed RLE + int left=width; + while (left>0) { + int count = stbi__get8(s), i; + if (stbi__at_eof(s)) return stbi__errpuc("bad file","file too short (mixed read count)"); + + if (count >= 128) { // Repeated + stbi_uc value[4]; + + if (count==128) + count = stbi__get16be(s); + else + count -= 127; + if (count > left) + return stbi__errpuc("bad file","scanline overrun"); + + if (!stbi__readval(s,packet->channel,value)) + return 0; + + for(i=0;ichannel,dest,value); + } else { // Raw + ++count; + if (count>left) return stbi__errpuc("bad file","scanline overrun"); + + for(i=0;ichannel,dest)) + return 0; + } + left-=count; + } + break; + } + } + } + } + + return result; +} + +static void *stbi__pic_load(stbi__context *s,int *px,int *py,int *comp,int req_comp, stbi__result_info *ri) +{ + stbi_uc *result; + int i, x,y, internal_comp; + STBI_NOTUSED(ri); + + if (!comp) comp = &internal_comp; + + for (i=0; i<92; ++i) + stbi__get8(s); + + x = stbi__get16be(s); + y = stbi__get16be(s); + if (stbi__at_eof(s)) return stbi__errpuc("bad file","file too short (pic header)"); + if (!stbi__mad3sizes_valid(x, y, 4, 0)) return stbi__errpuc("too large", "PIC image too large to decode"); + + stbi__get32be(s); //skip `ratio' + stbi__get16be(s); //skip `fields' + stbi__get16be(s); //skip `pad' + + // intermediate buffer is RGBA + result = (stbi_uc *) stbi__malloc_mad3(x, y, 4, 0); + memset(result, 0xff, x*y*4); + + if (!stbi__pic_load_core(s,x,y,comp, result)) { + STBI_FREE(result); + result=0; + } + *px = x; + *py = y; + if (req_comp == 0) req_comp = *comp; + result=stbi__convert_format(result,4,req_comp,x,y); + + return result; +} + +static int stbi__pic_test(stbi__context *s) +{ + int r = stbi__pic_test_core(s); + stbi__rewind(s); + return r; +} +#endif + +// ************************************************************************************************* +// GIF loader -- public domain by Jean-Marc Lienher -- simplified/shrunk by stb + +#ifndef STBI_NO_GIF +typedef struct +{ + stbi__int16 prefix; + stbi_uc first; + stbi_uc suffix; +} stbi__gif_lzw; + +typedef struct +{ + int w,h; + stbi_uc *out; // output buffer (always 4 components) + stbi_uc *background; // The current "background" as far as a gif is concerned + stbi_uc *history; + int flags, bgindex, ratio, transparent, eflags; + stbi_uc pal[256][4]; + stbi_uc lpal[256][4]; + stbi__gif_lzw codes[8192]; + stbi_uc *color_table; + int parse, step; + int lflags; + int start_x, start_y; + int max_x, max_y; + int cur_x, cur_y; + int line_size; + int delay; +} stbi__gif; + +static int stbi__gif_test_raw(stbi__context *s) +{ + int sz; + if (stbi__get8(s) != 'G' || stbi__get8(s) != 'I' || stbi__get8(s) != 'F' || stbi__get8(s) != '8') return 0; + sz = stbi__get8(s); + if (sz != '9' && sz != '7') return 0; + if (stbi__get8(s) != 'a') return 0; + return 1; +} + +static int stbi__gif_test(stbi__context *s) +{ + int r = stbi__gif_test_raw(s); + stbi__rewind(s); + return r; +} + +static void stbi__gif_parse_colortable(stbi__context *s, stbi_uc pal[256][4], int num_entries, int transp) +{ + int i; + for (i=0; i < num_entries; ++i) { + pal[i][2] = stbi__get8(s); + pal[i][1] = stbi__get8(s); + pal[i][0] = stbi__get8(s); + pal[i][3] = transp == i ? 0 : 255; + } +} + +static int stbi__gif_header(stbi__context *s, stbi__gif *g, int *comp, int is_info) +{ + stbi_uc version; + if (stbi__get8(s) != 'G' || stbi__get8(s) != 'I' || stbi__get8(s) != 'F' || stbi__get8(s) != '8') + return stbi__err("not GIF", "Corrupt GIF"); + + version = stbi__get8(s); + if (version != '7' && version != '9') return stbi__err("not GIF", "Corrupt GIF"); + if (stbi__get8(s) != 'a') return stbi__err("not GIF", "Corrupt GIF"); + + stbi__g_failure_reason = ""; + g->w = stbi__get16le(s); + g->h = stbi__get16le(s); + g->flags = stbi__get8(s); + g->bgindex = stbi__get8(s); + g->ratio = stbi__get8(s); + g->transparent = -1; + + if (comp != 0) *comp = 4; // can't actually tell whether it's 3 or 4 until we parse the comments + + if (is_info) return 1; + + if (g->flags & 0x80) + stbi__gif_parse_colortable(s,g->pal, 2 << (g->flags & 7), -1); + + return 1; +} + +static int stbi__gif_info_raw(stbi__context *s, int *x, int *y, int *comp) +{ + stbi__gif* g = (stbi__gif*) stbi__malloc(sizeof(stbi__gif)); + if (!stbi__gif_header(s, g, comp, 1)) { + STBI_FREE(g); + stbi__rewind( s ); + return 0; + } + if (x) *x = g->w; + if (y) *y = g->h; + STBI_FREE(g); + return 1; +} + +static void stbi__out_gif_code(stbi__gif *g, stbi__uint16 code) +{ + stbi_uc *p, *c; + int idx; + + // recurse to decode the prefixes, since the linked-list is backwards, + // and working backwards through an interleaved image would be nasty + if (g->codes[code].prefix >= 0) + stbi__out_gif_code(g, g->codes[code].prefix); + + if (g->cur_y >= g->max_y) return; + + idx = g->cur_x + g->cur_y; + p = &g->out[idx]; + g->history[idx / 4] = 1; + + c = &g->color_table[g->codes[code].suffix * 4]; + if (c[3] > 128) { // don't render transparent pixels; + p[0] = c[2]; + p[1] = c[1]; + p[2] = c[0]; + p[3] = c[3]; + } + g->cur_x += 4; + + if (g->cur_x >= g->max_x) { + g->cur_x = g->start_x; + g->cur_y += g->step; + + while (g->cur_y >= g->max_y && g->parse > 0) { + g->step = (1 << g->parse) * g->line_size; + g->cur_y = g->start_y + (g->step >> 1); + --g->parse; + } + } +} + +static stbi_uc *stbi__process_gif_raster(stbi__context *s, stbi__gif *g) +{ + stbi_uc lzw_cs; + stbi__int32 len, init_code; + stbi__uint32 first; + stbi__int32 codesize, codemask, avail, oldcode, bits, valid_bits, clear; + stbi__gif_lzw *p; + + lzw_cs = stbi__get8(s); + if (lzw_cs > 12) return NULL; + clear = 1 << lzw_cs; + first = 1; + codesize = lzw_cs + 1; + codemask = (1 << codesize) - 1; + bits = 0; + valid_bits = 0; + for (init_code = 0; init_code < clear; init_code++) { + g->codes[init_code].prefix = -1; + g->codes[init_code].first = (stbi_uc) init_code; + g->codes[init_code].suffix = (stbi_uc) init_code; + } + + // support no starting clear code + avail = clear+2; + oldcode = -1; + + len = 0; + for(;;) { + if (valid_bits < codesize) { + if (len == 0) { + len = stbi__get8(s); // start new block + if (len == 0) + return g->out; + } + --len; + bits |= (stbi__int32) stbi__get8(s) << valid_bits; + valid_bits += 8; + } else { + stbi__int32 code = bits & codemask; + bits >>= codesize; + valid_bits -= codesize; + // @OPTIMIZE: is there some way we can accelerate the non-clear path? + if (code == clear) { // clear code + codesize = lzw_cs + 1; + codemask = (1 << codesize) - 1; + avail = clear + 2; + oldcode = -1; + first = 0; + } else if (code == clear + 1) { // end of stream code + stbi__skip(s, len); + while ((len = stbi__get8(s)) > 0) + stbi__skip(s,len); + return g->out; + } else if (code <= avail) { + if (first) { + return stbi__errpuc("no clear code", "Corrupt GIF"); + } + + if (oldcode >= 0) { + p = &g->codes[avail++]; + if (avail > 8192) { + return stbi__errpuc("too many codes", "Corrupt GIF"); + } + + p->prefix = (stbi__int16) oldcode; + p->first = g->codes[oldcode].first; + p->suffix = (code == avail) ? p->first : g->codes[code].first; + } else if (code == avail) + return stbi__errpuc("illegal code in raster", "Corrupt GIF"); + + stbi__out_gif_code(g, (stbi__uint16) code); + + if ((avail & codemask) == 0 && avail <= 0x0FFF) { + codesize++; + codemask = (1 << codesize) - 1; + } + + oldcode = code; + } else { + return stbi__errpuc("illegal code in raster", "Corrupt GIF"); + } + } + } +} + +// this function is designed to support animated gifs, although stb_image doesn't support it +// two back is the image from two frames ago, used for a very specific disposal format +static stbi_uc *stbi__gif_load_next(stbi__context *s, stbi__gif *g, int *comp, int req_comp, stbi_uc *two_back) +{ + int dispose; + int first_frame; + int pi; + int pcount; + STBI_NOTUSED(req_comp); + + // on first frame, any non-written pixels get the background colour (non-transparent) + first_frame = 0; + if (g->out == 0) { + if (!stbi__gif_header(s, g, comp,0)) return 0; // stbi__g_failure_reason set by stbi__gif_header + if (!stbi__mad3sizes_valid(4, g->w, g->h, 0)) + return stbi__errpuc("too large", "GIF image is too large"); + pcount = g->w * g->h; + g->out = (stbi_uc *) stbi__malloc(4 * pcount); + g->background = (stbi_uc *) stbi__malloc(4 * pcount); + g->history = (stbi_uc *) stbi__malloc(pcount); + if (!g->out || !g->background || !g->history) + return stbi__errpuc("outofmem", "Out of memory"); + + // image is treated as "transparent" at the start - ie, nothing overwrites the current background; + // background colour is only used for pixels that are not rendered first frame, after that "background" + // color refers to the color that was there the previous frame. + memset(g->out, 0x00, 4 * pcount); + memset(g->background, 0x00, 4 * pcount); // state of the background (starts transparent) + memset(g->history, 0x00, pcount); // pixels that were affected previous frame + first_frame = 1; + } else { + // second frame - how do we dispoase of the previous one? + dispose = (g->eflags & 0x1C) >> 2; + pcount = g->w * g->h; + + if ((dispose == 3) && (two_back == 0)) { + dispose = 2; // if I don't have an image to revert back to, default to the old background + } + + if (dispose == 3) { // use previous graphic + for (pi = 0; pi < pcount; ++pi) { + if (g->history[pi]) { + memcpy( &g->out[pi * 4], &two_back[pi * 4], 4 ); + } + } + } else if (dispose == 2) { + // restore what was changed last frame to background before that frame; + for (pi = 0; pi < pcount; ++pi) { + if (g->history[pi]) { + memcpy( &g->out[pi * 4], &g->background[pi * 4], 4 ); + } + } + } else { + // This is a non-disposal case eithe way, so just + // leave the pixels as is, and they will become the new background + // 1: do not dispose + // 0: not specified. + } + + // background is what out is after the undoing of the previou frame; + memcpy( g->background, g->out, 4 * g->w * g->h ); + } + + // clear my history; + memset( g->history, 0x00, g->w * g->h ); // pixels that were affected previous frame + + for (;;) { + int tag = stbi__get8(s); + switch (tag) { + case 0x2C: /* Image Descriptor */ + { + stbi__int32 x, y, w, h; + stbi_uc *o; + + x = stbi__get16le(s); + y = stbi__get16le(s); + w = stbi__get16le(s); + h = stbi__get16le(s); + if (((x + w) > (g->w)) || ((y + h) > (g->h))) + return stbi__errpuc("bad Image Descriptor", "Corrupt GIF"); + + g->line_size = g->w * 4; + g->start_x = x * 4; + g->start_y = y * g->line_size; + g->max_x = g->start_x + w * 4; + g->max_y = g->start_y + h * g->line_size; + g->cur_x = g->start_x; + g->cur_y = g->start_y; + + // if the width of the specified rectangle is 0, that means + // we may not see *any* pixels or the image is malformed; + // to make sure this is caught, move the current y down to + // max_y (which is what out_gif_code checks). + if (w == 0) + g->cur_y = g->max_y; + + g->lflags = stbi__get8(s); + + if (g->lflags & 0x40) { + g->step = 8 * g->line_size; // first interlaced spacing + g->parse = 3; + } else { + g->step = g->line_size; + g->parse = 0; + } + + if (g->lflags & 0x80) { + stbi__gif_parse_colortable(s,g->lpal, 2 << (g->lflags & 7), g->eflags & 0x01 ? g->transparent : -1); + g->color_table = (stbi_uc *) g->lpal; + } else if (g->flags & 0x80) { + g->color_table = (stbi_uc *) g->pal; + } else + return stbi__errpuc("missing color table", "Corrupt GIF"); + + o = stbi__process_gif_raster(s, g); + if (!o) return NULL; + + // if this was the first frame, + pcount = g->w * g->h; + if (first_frame && (g->bgindex > 0)) { + // if first frame, any pixel not drawn to gets the background color + for (pi = 0; pi < pcount; ++pi) { + if (g->history[pi] == 0) { + g->pal[g->bgindex][3] = 255; // just in case it was made transparent, undo that; It will be reset next frame if need be; + memcpy( &g->out[pi * 4], &g->pal[g->bgindex], 4 ); + } + } + } + + return o; + } + + case 0x21: // Comment Extension. + { + int len; + int ext = stbi__get8(s); + if (ext == 0xF9) { // Graphic Control Extension. + len = stbi__get8(s); + if (len == 4) { + g->eflags = stbi__get8(s); + g->delay = 10 * stbi__get16le(s); // delay - 1/100th of a second, saving as 1/1000ths. + + // unset old transparent + if (g->transparent >= 0) { + g->pal[g->transparent][3] = 255; + } + if (g->eflags & 0x01) { + g->transparent = stbi__get8(s); + if (g->transparent >= 0) { + g->pal[g->transparent][3] = 0; + } + } else { + // don't need transparent + stbi__skip(s, 1); + g->transparent = -1; + } + } else { + stbi__skip(s, len); + break; + } + } + while ((len = stbi__get8(s)) != 0) { + stbi__skip(s, len); + } + break; + } + + case 0x3B: // gif stream termination code + return (stbi_uc *) s; // using '1' causes warning on some compilers + + default: + return stbi__errpuc("unknown code", "Corrupt GIF"); + } + } +} + +static void *stbi__load_gif_main(stbi__context *s, int **delays, int *x, int *y, int *z, int *comp, int req_comp) +{ + if (stbi__gif_test(s)) { + int layers = 0; + stbi_uc *u = 0; + stbi_uc *out = 0; + stbi_uc *two_back = 0; + stbi__gif g; + int stride; + memset(&g, 0, sizeof(g)); + if (delays) { + *delays = 0; + } + + do { + u = stbi__gif_load_next(s, &g, comp, req_comp, two_back); + if (u == (stbi_uc *) s) u = 0; // end of animated gif marker + + if (u) { + *x = g.w; + *y = g.h; + ++layers; + stride = g.w * g.h * 4; + + if (out) { + void *tmp = (stbi_uc*) STBI_REALLOC( out, layers * stride ); + if (NULL == tmp) { + STBI_FREE(g.out); + STBI_FREE(g.history); + STBI_FREE(g.background); + return stbi__errpuc("outofmem", "Out of memory"); + } + else + out = (stbi_uc*) tmp; + if (delays) { + *delays = (int*) STBI_REALLOC( *delays, sizeof(int) * layers ); + } + } else { + out = (stbi_uc*)stbi__malloc( layers * stride ); + if (delays) { + *delays = (int*) stbi__malloc( layers * sizeof(int) ); + } + } + memcpy( out + ((layers - 1) * stride), u, stride ); + if (layers >= 2) { + two_back = out - 2 * stride; + } + + if (delays) { + (*delays)[layers - 1U] = g.delay; + } + } + } while (u != 0); + + // free temp buffer; + STBI_FREE(g.out); + STBI_FREE(g.history); + STBI_FREE(g.background); + + // do the final conversion after loading everything; + if (req_comp && req_comp != 4) + out = stbi__convert_format(out, 4, req_comp, layers * g.w, g.h); + + *z = layers; + return out; + } else { + return stbi__errpuc("not GIF", "Image was not as a gif type."); + } +} + +static void *stbi__gif_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri) +{ + stbi_uc *u = 0; + stbi__gif g; + memset(&g, 0, sizeof(g)); + STBI_NOTUSED(ri); + + u = stbi__gif_load_next(s, &g, comp, req_comp, 0); + if (u == (stbi_uc *) s) u = 0; // end of animated gif marker + if (u) { + *x = g.w; + *y = g.h; + + // moved conversion to after successful load so that the same + // can be done for multiple frames. + if (req_comp && req_comp != 4) + u = stbi__convert_format(u, 4, req_comp, g.w, g.h); + } else if (g.out) { + // if there was an error and we allocated an image buffer, free it! + STBI_FREE(g.out); + } + + // free buffers needed for multiple frame loading; + STBI_FREE(g.history); + STBI_FREE(g.background); + + return u; +} + +static int stbi__gif_info(stbi__context *s, int *x, int *y, int *comp) +{ + return stbi__gif_info_raw(s,x,y,comp); +} +#endif + +// ************************************************************************************************* +// Radiance RGBE HDR loader +// originally by Nicolas Schulz +#ifndef STBI_NO_HDR +static int stbi__hdr_test_core(stbi__context *s, const char *signature) +{ + int i; + for (i=0; signature[i]; ++i) + if (stbi__get8(s) != signature[i]) + return 0; + stbi__rewind(s); + return 1; +} + +static int stbi__hdr_test(stbi__context* s) +{ + int r = stbi__hdr_test_core(s, "#?RADIANCE\n"); + stbi__rewind(s); + if(!r) { + r = stbi__hdr_test_core(s, "#?RGBE\n"); + stbi__rewind(s); + } + return r; +} + +#define STBI__HDR_BUFLEN 1024 +static char *stbi__hdr_gettoken(stbi__context *z, char *buffer) +{ + int len=0; + char c = '\0'; + + c = (char) stbi__get8(z); + + while (!stbi__at_eof(z) && c != '\n') { + buffer[len++] = c; + if (len == STBI__HDR_BUFLEN-1) { + // flush to end of line + while (!stbi__at_eof(z) && stbi__get8(z) != '\n') + ; + break; + } + c = (char) stbi__get8(z); + } + + buffer[len] = 0; + return buffer; +} + +static void stbi__hdr_convert(float *output, stbi_uc *input, int req_comp) +{ + if ( input[3] != 0 ) { + float f1; + // Exponent + f1 = (float) ldexp(1.0f, input[3] - (int)(128 + 8)); + if (req_comp <= 2) + output[0] = (input[0] + input[1] + input[2]) * f1 / 3; + else { + output[0] = input[0] * f1; + output[1] = input[1] * f1; + output[2] = input[2] * f1; + } + if (req_comp == 2) output[1] = 1; + if (req_comp == 4) output[3] = 1; + } else { + switch (req_comp) { + case 4: output[3] = 1; /* fallthrough */ + case 3: output[0] = output[1] = output[2] = 0; + break; + case 2: output[1] = 1; /* fallthrough */ + case 1: output[0] = 0; + break; + } + } +} + +static float *stbi__hdr_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri) +{ + char buffer[STBI__HDR_BUFLEN]; + char *token; + int valid = 0; + int width, height; + stbi_uc *scanline; + float *hdr_data; + int len; + unsigned char count, value; + int i, j, k, c1,c2, z; + const char *headerToken; + STBI_NOTUSED(ri); + + // Check identifier + headerToken = stbi__hdr_gettoken(s,buffer); + if (strcmp(headerToken, "#?RADIANCE") != 0 && strcmp(headerToken, "#?RGBE") != 0) + return stbi__errpf("not HDR", "Corrupt HDR image"); + + // Parse header + for(;;) { + token = stbi__hdr_gettoken(s,buffer); + if (token[0] == 0) break; + if (strcmp(token, "FORMAT=32-bit_rle_rgbe") == 0) valid = 1; + } + + if (!valid) return stbi__errpf("unsupported format", "Unsupported HDR format"); + + // Parse width and height + // can't use sscanf() if we're not using stdio! + token = stbi__hdr_gettoken(s,buffer); + if (strncmp(token, "-Y ", 3)) return stbi__errpf("unsupported data layout", "Unsupported HDR format"); + token += 3; + height = (int) strtol(token, &token, 10); + while (*token == ' ') ++token; + if (strncmp(token, "+X ", 3)) return stbi__errpf("unsupported data layout", "Unsupported HDR format"); + token += 3; + width = (int) strtol(token, NULL, 10); + + *x = width; + *y = height; + + if (comp) *comp = 3; + if (req_comp == 0) req_comp = 3; + + if (!stbi__mad4sizes_valid(width, height, req_comp, sizeof(float), 0)) + return stbi__errpf("too large", "HDR image is too large"); + + // Read data + hdr_data = (float *) stbi__malloc_mad4(width, height, req_comp, sizeof(float), 0); + if (!hdr_data) + return stbi__errpf("outofmem", "Out of memory"); + + // Load image data + // image data is stored as some number of sca + if ( width < 8 || width >= 32768) { + // Read flat data + for (j=0; j < height; ++j) { + for (i=0; i < width; ++i) { + stbi_uc rgbe[4]; + main_decode_loop: + stbi__getn(s, rgbe, 4); + stbi__hdr_convert(hdr_data + j * width * req_comp + i * req_comp, rgbe, req_comp); + } + } + } else { + // Read RLE-encoded data + scanline = NULL; + + for (j = 0; j < height; ++j) { + c1 = stbi__get8(s); + c2 = stbi__get8(s); + len = stbi__get8(s); + if (c1 != 2 || c2 != 2 || (len & 0x80)) { + // not run-length encoded, so we have to actually use THIS data as a decoded + // pixel (note this can't be a valid pixel--one of RGB must be >= 128) + stbi_uc rgbe[4]; + rgbe[0] = (stbi_uc) c1; + rgbe[1] = (stbi_uc) c2; + rgbe[2] = (stbi_uc) len; + rgbe[3] = (stbi_uc) stbi__get8(s); + stbi__hdr_convert(hdr_data, rgbe, req_comp); + i = 1; + j = 0; + STBI_FREE(scanline); + goto main_decode_loop; // yes, this makes no sense + } + len <<= 8; + len |= stbi__get8(s); + if (len != width) { STBI_FREE(hdr_data); STBI_FREE(scanline); return stbi__errpf("invalid decoded scanline length", "corrupt HDR"); } + if (scanline == NULL) { + scanline = (stbi_uc *) stbi__malloc_mad2(width, 4, 0); + if (!scanline) { + STBI_FREE(hdr_data); + return stbi__errpf("outofmem", "Out of memory"); + } + } + + for (k = 0; k < 4; ++k) { + int nleft; + i = 0; + while ((nleft = width - i) > 0) { + count = stbi__get8(s); + if (count > 128) { + // Run + value = stbi__get8(s); + count -= 128; + if (count > nleft) { STBI_FREE(hdr_data); STBI_FREE(scanline); return stbi__errpf("corrupt", "bad RLE data in HDR"); } + for (z = 0; z < count; ++z) + scanline[i++ * 4 + k] = value; + } else { + // Dump + if (count > nleft) { STBI_FREE(hdr_data); STBI_FREE(scanline); return stbi__errpf("corrupt", "bad RLE data in HDR"); } + for (z = 0; z < count; ++z) + scanline[i++ * 4 + k] = stbi__get8(s); + } + } + } + for (i=0; i < width; ++i) + stbi__hdr_convert(hdr_data+(j*width + i)*req_comp, scanline + i*4, req_comp); + } + if (scanline) + STBI_FREE(scanline); + } + + return hdr_data; +} + +static int stbi__hdr_info(stbi__context *s, int *x, int *y, int *comp) +{ + char buffer[STBI__HDR_BUFLEN]; + char *token; + int valid = 0; + int dummy; + + if (!x) x = &dummy; + if (!y) y = &dummy; + if (!comp) comp = &dummy; + + if (stbi__hdr_test(s) == 0) { + stbi__rewind( s ); + return 0; + } + + for(;;) { + token = stbi__hdr_gettoken(s,buffer); + if (token[0] == 0) break; + if (strcmp(token, "FORMAT=32-bit_rle_rgbe") == 0) valid = 1; + } + + if (!valid) { + stbi__rewind( s ); + return 0; + } + token = stbi__hdr_gettoken(s,buffer); + if (strncmp(token, "-Y ", 3)) { + stbi__rewind( s ); + return 0; + } + token += 3; + *y = (int) strtol(token, &token, 10); + while (*token == ' ') ++token; + if (strncmp(token, "+X ", 3)) { + stbi__rewind( s ); + return 0; + } + token += 3; + *x = (int) strtol(token, NULL, 10); + *comp = 3; + return 1; +} +#endif // STBI_NO_HDR + +#ifndef STBI_NO_BMP +static int stbi__bmp_info(stbi__context *s, int *x, int *y, int *comp) +{ + void *p; + stbi__bmp_data info; + + info.all_a = 255; + p = stbi__bmp_parse_header(s, &info); + stbi__rewind( s ); + if (p == NULL) + return 0; + if (x) *x = s->img_x; + if (y) *y = s->img_y; + if (comp) { + if (info.bpp == 24 && info.ma == 0xff000000) + *comp = 3; + else + *comp = info.ma ? 4 : 3; + } + return 1; +} +#endif + +#ifndef STBI_NO_PSD +static int stbi__psd_info(stbi__context *s, int *x, int *y, int *comp) +{ + int channelCount, dummy, depth; + if (!x) x = &dummy; + if (!y) y = &dummy; + if (!comp) comp = &dummy; + if (stbi__get32be(s) != 0x38425053) { + stbi__rewind( s ); + return 0; + } + if (stbi__get16be(s) != 1) { + stbi__rewind( s ); + return 0; + } + stbi__skip(s, 6); + channelCount = stbi__get16be(s); + if (channelCount < 0 || channelCount > 16) { + stbi__rewind( s ); + return 0; + } + *y = stbi__get32be(s); + *x = stbi__get32be(s); + depth = stbi__get16be(s); + if (depth != 8 && depth != 16) { + stbi__rewind( s ); + return 0; + } + if (stbi__get16be(s) != 3) { + stbi__rewind( s ); + return 0; + } + *comp = 4; + return 1; +} + +static int stbi__psd_is16(stbi__context *s) +{ + int channelCount, depth; + if (stbi__get32be(s) != 0x38425053) { + stbi__rewind( s ); + return 0; + } + if (stbi__get16be(s) != 1) { + stbi__rewind( s ); + return 0; + } + stbi__skip(s, 6); + channelCount = stbi__get16be(s); + if (channelCount < 0 || channelCount > 16) { + stbi__rewind( s ); + return 0; + } + (void) stbi__get32be(s); + (void) stbi__get32be(s); + depth = stbi__get16be(s); + if (depth != 16) { + stbi__rewind( s ); + return 0; + } + return 1; +} +#endif + +#ifndef STBI_NO_PIC +static int stbi__pic_info(stbi__context *s, int *x, int *y, int *comp) +{ + int act_comp=0,num_packets=0,chained,dummy; + stbi__pic_packet packets[10]; + + if (!x) x = &dummy; + if (!y) y = &dummy; + if (!comp) comp = &dummy; + + if (!stbi__pic_is4(s,"\x53\x80\xF6\x34")) { + stbi__rewind(s); + return 0; + } + + stbi__skip(s, 88); + + *x = stbi__get16be(s); + *y = stbi__get16be(s); + if (stbi__at_eof(s)) { + stbi__rewind( s); + return 0; + } + if ( (*x) != 0 && (1 << 28) / (*x) < (*y)) { + stbi__rewind( s ); + return 0; + } + + stbi__skip(s, 8); + + do { + stbi__pic_packet *packet; + + if (num_packets==sizeof(packets)/sizeof(packets[0])) + return 0; + + packet = &packets[num_packets++]; + chained = stbi__get8(s); + packet->size = stbi__get8(s); + packet->type = stbi__get8(s); + packet->channel = stbi__get8(s); + act_comp |= packet->channel; + + if (stbi__at_eof(s)) { + stbi__rewind( s ); + return 0; + } + if (packet->size != 8) { + stbi__rewind( s ); + return 0; + } + } while (chained); + + *comp = (act_comp & 0x10 ? 4 : 3); + + return 1; +} +#endif + +// ************************************************************************************************* +// Portable Gray Map and Portable Pixel Map loader +// by Ken Miller +// +// PGM: http://netpbm.sourceforge.net/doc/pgm.html +// PPM: http://netpbm.sourceforge.net/doc/ppm.html +// +// Known limitations: +// Does not support comments in the header section +// Does not support ASCII image data (formats P2 and P3) +// Does not support 16-bit-per-channel + +#ifndef STBI_NO_PNM + +static int stbi__pnm_test(stbi__context *s) +{ + char p, t; + p = (char) stbi__get8(s); + t = (char) stbi__get8(s); + if (p != 'P' || (t != '5' && t != '6')) { + stbi__rewind( s ); + return 0; + } + return 1; +} + +static void *stbi__pnm_load(stbi__context *s, int *x, int *y, int *comp, int req_comp, stbi__result_info *ri) +{ + stbi_uc *out; + STBI_NOTUSED(ri); + + if (!stbi__pnm_info(s, (int *)&s->img_x, (int *)&s->img_y, (int *)&s->img_n)) + return 0; + + *x = s->img_x; + *y = s->img_y; + if (comp) *comp = s->img_n; + + if (!stbi__mad3sizes_valid(s->img_n, s->img_x, s->img_y, 0)) + return stbi__errpuc("too large", "PNM too large"); + + out = (stbi_uc *) stbi__malloc_mad3(s->img_n, s->img_x, s->img_y, 0); + if (!out) return stbi__errpuc("outofmem", "Out of memory"); + stbi__getn(s, out, s->img_n * s->img_x * s->img_y); + + if (req_comp && req_comp != s->img_n) { + out = stbi__convert_format(out, s->img_n, req_comp, s->img_x, s->img_y); + if (out == NULL) return out; // stbi__convert_format frees input on failure + } + return out; +} + +static int stbi__pnm_isspace(char c) +{ + return c == ' ' || c == '\t' || c == '\n' || c == '\v' || c == '\f' || c == '\r'; +} + +static void stbi__pnm_skip_whitespace(stbi__context *s, char *c) +{ + for (;;) { + while (!stbi__at_eof(s) && stbi__pnm_isspace(*c)) + *c = (char) stbi__get8(s); + + if (stbi__at_eof(s) || *c != '#') + break; + + while (!stbi__at_eof(s) && *c != '\n' && *c != '\r' ) + *c = (char) stbi__get8(s); + } +} + +static int stbi__pnm_isdigit(char c) +{ + return c >= '0' && c <= '9'; +} + +static int stbi__pnm_getinteger(stbi__context *s, char *c) +{ + int value = 0; + + while (!stbi__at_eof(s) && stbi__pnm_isdigit(*c)) { + value = value*10 + (*c - '0'); + *c = (char) stbi__get8(s); + } + + return value; +} + +static int stbi__pnm_info(stbi__context *s, int *x, int *y, int *comp) +{ + int maxv, dummy; + char c, p, t; + + if (!x) x = &dummy; + if (!y) y = &dummy; + if (!comp) comp = &dummy; + + stbi__rewind(s); + + // Get identifier + p = (char) stbi__get8(s); + t = (char) stbi__get8(s); + if (p != 'P' || (t != '5' && t != '6')) { + stbi__rewind(s); + return 0; + } + + *comp = (t == '6') ? 3 : 1; // '5' is 1-component .pgm; '6' is 3-component .ppm + + c = (char) stbi__get8(s); + stbi__pnm_skip_whitespace(s, &c); + + *x = stbi__pnm_getinteger(s, &c); // read width + stbi__pnm_skip_whitespace(s, &c); + + *y = stbi__pnm_getinteger(s, &c); // read height + stbi__pnm_skip_whitespace(s, &c); + + maxv = stbi__pnm_getinteger(s, &c); // read max value + + if (maxv > 255) + return stbi__err("max value > 255", "PPM image not 8-bit"); + else + return 1; +} +#endif + +static int stbi__info_main(stbi__context *s, int *x, int *y, int *comp) +{ + #ifndef STBI_NO_JPEG + if (stbi__jpeg_info(s, x, y, comp)) return 1; + #endif + + #ifndef STBI_NO_PNG + if (stbi__png_info(s, x, y, comp)) return 1; + #endif + + #ifndef STBI_NO_GIF + if (stbi__gif_info(s, x, y, comp)) return 1; + #endif + + #ifndef STBI_NO_BMP + if (stbi__bmp_info(s, x, y, comp)) return 1; + #endif + + #ifndef STBI_NO_PSD + if (stbi__psd_info(s, x, y, comp)) return 1; + #endif + + #ifndef STBI_NO_PIC + if (stbi__pic_info(s, x, y, comp)) return 1; + #endif + + #ifndef STBI_NO_PNM + if (stbi__pnm_info(s, x, y, comp)) return 1; + #endif + + #ifndef STBI_NO_HDR + if (stbi__hdr_info(s, x, y, comp)) return 1; + #endif + + // test tga last because it's a crappy test! + #ifndef STBI_NO_TGA + if (stbi__tga_info(s, x, y, comp)) + return 1; + #endif + return stbi__err("unknown image type", "Image not of any known type, or corrupt"); +} + +static int stbi__is_16_main(stbi__context *s) +{ + #ifndef STBI_NO_PNG + if (stbi__png_is16(s)) return 1; + #endif + + #ifndef STBI_NO_PSD + if (stbi__psd_is16(s)) return 1; + #endif + + return 0; +} + +#ifndef STBI_NO_STDIO +STBIDEF int stbi_info(char const *filename, int *x, int *y, int *comp) +{ + FILE *f = stbi__fopen(filename, "rb"); + int result; + if (!f) return stbi__err("can't fopen", "Unable to open file"); + result = stbi_info_from_file(f, x, y, comp); + fclose(f); + return result; +} + +STBIDEF int stbi_info_from_file(FILE *f, int *x, int *y, int *comp) +{ + int r; + stbi__context s; + long pos = ftell(f); + stbi__start_file(&s, f); + r = stbi__info_main(&s,x,y,comp); + fseek(f,pos,SEEK_SET); + return r; +} + +STBIDEF int stbi_is_16_bit(char const *filename) +{ + FILE *f = stbi__fopen(filename, "rb"); + int result; + if (!f) return stbi__err("can't fopen", "Unable to open file"); + result = stbi_is_16_bit_from_file(f); + fclose(f); + return result; +} + +STBIDEF int stbi_is_16_bit_from_file(FILE *f) +{ + int r; + stbi__context s; + long pos = ftell(f); + stbi__start_file(&s, f); + r = stbi__is_16_main(&s); + fseek(f,pos,SEEK_SET); + return r; +} +#endif // !STBI_NO_STDIO + +STBIDEF int stbi_info_from_memory(stbi_uc const *buffer, int len, int *x, int *y, int *comp) +{ + stbi__context s; + stbi__start_mem(&s,buffer,len); + return stbi__info_main(&s,x,y,comp); +} + +STBIDEF int stbi_info_from_callbacks(stbi_io_callbacks const *c, void *user, int *x, int *y, int *comp) +{ + stbi__context s; + stbi__start_callbacks(&s, (stbi_io_callbacks *) c, user); + return stbi__info_main(&s,x,y,comp); +} + +STBIDEF int stbi_is_16_bit_from_memory(stbi_uc const *buffer, int len) +{ + stbi__context s; + stbi__start_mem(&s,buffer,len); + return stbi__is_16_main(&s); +} + +STBIDEF int stbi_is_16_bit_from_callbacks(stbi_io_callbacks const *c, void *user) +{ + stbi__context s; + stbi__start_callbacks(&s, (stbi_io_callbacks *) c, user); + return stbi__is_16_main(&s); +} + +#endif // STB_IMAGE_IMPLEMENTATION + +/* + revision history: + 2.20 (2019-02-07) support utf8 filenames in Windows; fix warnings and platform ifdefs + 2.19 (2018-02-11) fix warning + 2.18 (2018-01-30) fix warnings + 2.17 (2018-01-29) change sbti__shiftsigned to avoid clang -O2 bug + 1-bit BMP + *_is_16_bit api + avoid warnings + 2.16 (2017-07-23) all functions have 16-bit variants; + STBI_NO_STDIO works again; + compilation fixes; + fix rounding in unpremultiply; + optimize vertical flip; + disable raw_len validation; + documentation fixes + 2.15 (2017-03-18) fix png-1,2,4 bug; now all Imagenet JPGs decode; + warning fixes; disable run-time SSE detection on gcc; + uniform handling of optional "return" values; + thread-safe initialization of zlib tables + 2.14 (2017-03-03) remove deprecated STBI_JPEG_OLD; fixes for Imagenet JPGs + 2.13 (2016-11-29) add 16-bit API, only supported for PNG right now + 2.12 (2016-04-02) fix typo in 2.11 PSD fix that caused crashes + 2.11 (2016-04-02) allocate large structures on the stack + remove white matting for transparent PSD + fix reported channel count for PNG & BMP + re-enable SSE2 in non-gcc 64-bit + support RGB-formatted JPEG + read 16-bit PNGs (only as 8-bit) + 2.10 (2016-01-22) avoid warning introduced in 2.09 by STBI_REALLOC_SIZED + 2.09 (2016-01-16) allow comments in PNM files + 16-bit-per-pixel TGA (not bit-per-component) + info() for TGA could break due to .hdr handling + info() for BMP to shares code instead of sloppy parse + can use STBI_REALLOC_SIZED if allocator doesn't support realloc + code cleanup + 2.08 (2015-09-13) fix to 2.07 cleanup, reading RGB PSD as RGBA + 2.07 (2015-09-13) fix compiler warnings + partial animated GIF support + limited 16-bpc PSD support + #ifdef unused functions + bug with < 92 byte PIC,PNM,HDR,TGA + 2.06 (2015-04-19) fix bug where PSD returns wrong '*comp' value + 2.05 (2015-04-19) fix bug in progressive JPEG handling, fix warning + 2.04 (2015-04-15) try to re-enable SIMD on MinGW 64-bit + 2.03 (2015-04-12) extra corruption checking (mmozeiko) + stbi_set_flip_vertically_on_load (nguillemot) + fix NEON support; fix mingw support + 2.02 (2015-01-19) fix incorrect assert, fix warning + 2.01 (2015-01-17) fix various warnings; suppress SIMD on gcc 32-bit without -msse2 + 2.00b (2014-12-25) fix STBI_MALLOC in progressive JPEG + 2.00 (2014-12-25) optimize JPG, including x86 SSE2 & NEON SIMD (ryg) + progressive JPEG (stb) + PGM/PPM support (Ken Miller) + STBI_MALLOC,STBI_REALLOC,STBI_FREE + GIF bugfix -- seemingly never worked + STBI_NO_*, STBI_ONLY_* + 1.48 (2014-12-14) fix incorrectly-named assert() + 1.47 (2014-12-14) 1/2/4-bit PNG support, both direct and paletted (Omar Cornut & stb) + optimize PNG (ryg) + fix bug in interlaced PNG with user-specified channel count (stb) + 1.46 (2014-08-26) + fix broken tRNS chunk (colorkey-style transparency) in non-paletted PNG + 1.45 (2014-08-16) + fix MSVC-ARM internal compiler error by wrapping malloc + 1.44 (2014-08-07) + various warning fixes from Ronny Chevalier + 1.43 (2014-07-15) + fix MSVC-only compiler problem in code changed in 1.42 + 1.42 (2014-07-09) + don't define _CRT_SECURE_NO_WARNINGS (affects user code) + fixes to stbi__cleanup_jpeg path + added STBI_ASSERT to avoid requiring assert.h + 1.41 (2014-06-25) + fix search&replace from 1.36 that messed up comments/error messages + 1.40 (2014-06-22) + fix gcc struct-initialization warning + 1.39 (2014-06-15) + fix to TGA optimization when req_comp != number of components in TGA; + fix to GIF loading because BMP wasn't rewinding (whoops, no GIFs in my test suite) + add support for BMP version 5 (more ignored fields) + 1.38 (2014-06-06) + suppress MSVC warnings on integer casts truncating values + fix accidental rename of 'skip' field of I/O + 1.37 (2014-06-04) + remove duplicate typedef + 1.36 (2014-06-03) + convert to header file single-file library + if de-iphone isn't set, load iphone images color-swapped instead of returning NULL + 1.35 (2014-05-27) + various warnings + fix broken STBI_SIMD path + fix bug where stbi_load_from_file no longer left file pointer in correct place + fix broken non-easy path for 32-bit BMP (possibly never used) + TGA optimization by Arseny Kapoulkine + 1.34 (unknown) + use STBI_NOTUSED in stbi__resample_row_generic(), fix one more leak in tga failure case + 1.33 (2011-07-14) + make stbi_is_hdr work in STBI_NO_HDR (as specified), minor compiler-friendly improvements + 1.32 (2011-07-13) + support for "info" function for all supported filetypes (SpartanJ) + 1.31 (2011-06-20) + a few more leak fixes, bug in PNG handling (SpartanJ) + 1.30 (2011-06-11) + added ability to load files via callbacks to accomidate custom input streams (Ben Wenger) + removed deprecated format-specific test/load functions + removed support for installable file formats (stbi_loader) -- would have been broken for IO callbacks anyway + error cases in bmp and tga give messages and don't leak (Raymond Barbiero, grisha) + fix inefficiency in decoding 32-bit BMP (David Woo) + 1.29 (2010-08-16) + various warning fixes from Aurelien Pocheville + 1.28 (2010-08-01) + fix bug in GIF palette transparency (SpartanJ) + 1.27 (2010-08-01) + cast-to-stbi_uc to fix warnings + 1.26 (2010-07-24) + fix bug in file buffering for PNG reported by SpartanJ + 1.25 (2010-07-17) + refix trans_data warning (Won Chun) + 1.24 (2010-07-12) + perf improvements reading from files on platforms with lock-heavy fgetc() + minor perf improvements for jpeg + deprecated type-specific functions so we'll get feedback if they're needed + attempt to fix trans_data warning (Won Chun) + 1.23 fixed bug in iPhone support + 1.22 (2010-07-10) + removed image *writing* support + stbi_info support from Jetro Lauha + GIF support from Jean-Marc Lienher + iPhone PNG-extensions from James Brown + warning-fixes from Nicolas Schulz and Janez Zemva (i.stbi__err. Janez (U+017D)emva) + 1.21 fix use of 'stbi_uc' in header (reported by jon blow) + 1.20 added support for Softimage PIC, by Tom Seddon + 1.19 bug in interlaced PNG corruption check (found by ryg) + 1.18 (2008-08-02) + fix a threading bug (local mutable static) + 1.17 support interlaced PNG + 1.16 major bugfix - stbi__convert_format converted one too many pixels + 1.15 initialize some fields for thread safety + 1.14 fix threadsafe conversion bug + header-file-only version (#define STBI_HEADER_FILE_ONLY before including) + 1.13 threadsafe + 1.12 const qualifiers in the API + 1.11 Support installable IDCT, colorspace conversion routines + 1.10 Fixes for 64-bit (don't use "unsigned long") + optimized upsampling by Fabian "ryg" Giesen + 1.09 Fix format-conversion for PSD code (bad global variables!) + 1.08 Thatcher Ulrich's PSD code integrated by Nicolas Schulz + 1.07 attempt to fix C++ warning/errors again + 1.06 attempt to fix C++ warning/errors again + 1.05 fix TGA loading to return correct *comp and use good luminance calc + 1.04 default float alpha is 1, not 255; use 'void *' for stbi_image_free + 1.03 bugfixes to STBI_NO_STDIO, STBI_NO_HDR + 1.02 support for (subset of) HDR files, float interface for preferred access to them + 1.01 fix bug: possible bug in handling right-side up bmps... not sure + fix bug: the stbi__bmp_load() and stbi__tga_load() functions didn't work at all + 1.00 interface to zlib that skips zlib header + 0.99 correct handling of alpha in palette + 0.98 TGA loader by lonesock; dynamically add loaders (untested) + 0.97 jpeg errors on too large a file; also catch another malloc failure + 0.96 fix detection of invalid v value - particleman@mollyrocket forum + 0.95 during header scan, seek to markers in case of padding + 0.94 STBI_NO_STDIO to disable stdio usage; rename all #defines the same + 0.93 handle jpegtran output; verbose errors + 0.92 read 4,8,16,24,32-bit BMP files of several formats + 0.91 output 24-bit Windows 3.0 BMP files + 0.90 fix a few more warnings; bump version number to approach 1.0 + 0.61 bugfixes due to Marc LeBlanc, Christopher Lloyd + 0.60 fix compiling as c++ + 0.59 fix warnings: merge Dave Moore's -Wall fixes + 0.58 fix bug: zlib uncompressed mode len/nlen was wrong endian + 0.57 fix bug: jpg last huffman symbol before marker was >9 bits but less than 16 available + 0.56 fix bug: zlib uncompressed mode len vs. nlen + 0.55 fix bug: restart_interval not initialized to 0 + 0.54 allow NULL for 'int *comp' + 0.53 fix bug in png 3->4; speedup png decoding + 0.52 png handles req_comp=3,4 directly; minor cleanup; jpeg comments + 0.51 obey req_comp requests, 1-component jpegs return as 1-component, + on 'test' only check type, not whether we support this variant + 0.50 (2006-11-19) + first released version +*/ + + +/* +------------------------------------------------------------------------------ +This software is available under 2 licenses -- choose whichever you prefer. +------------------------------------------------------------------------------ +ALTERNATIVE A - MIT License +Copyright (c) 2017 Sean Barrett +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. +------------------------------------------------------------------------------ +ALTERNATIVE B - Public Domain (www.unlicense.org) +This is free and unencumbered software released into the public domain. +Anyone is free to copy, modify, publish, use, compile, sell, or distribute this +software, either in source code form or as a compiled binary, for any purpose, +commercial or non-commercial, and by any means. +In jurisdictions that recognize copyright laws, the author or authors of this +software dedicate any and all copyright interest in the software to the public +domain. We make this dedication for the benefit of the public at large and to +the detriment of our heirs and successors. We intend this dedication to be an +overt act of relinquishment in perpetuity of all present and future rights to +this software under copyright law. +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 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. +------------------------------------------------------------------------------ +*/ diff --git a/private/blah/third_party/stb_image_write.h b/private/blah/third_party/stb_image_write.h new file mode 100644 index 0000000..488d50c --- /dev/null +++ b/private/blah/third_party/stb_image_write.h @@ -0,0 +1,1666 @@ +/* stb_image_write - v1.14 - public domain - http://nothings.org/stb + writes out PNG/BMP/TGA/JPEG/HDR images to C stdio - Sean Barrett 2010-2015 + no warranty implied; use at your own risk + + Before #including, + + #define STB_IMAGE_WRITE_IMPLEMENTATION + + in the file that you want to have the implementation. + + Will probably not work correctly with strict-aliasing optimizations. + +ABOUT: + + This header file is a library for writing images to C stdio or a callback. + + The PNG output is not optimal; it is 20-50% larger than the file + written by a decent optimizing implementation; though providing a custom + zlib compress function (see STBIW_ZLIB_COMPRESS) can mitigate that. + This library is designed for source code compactness and simplicity, + not optimal image file size or run-time performance. + +BUILDING: + + You can #define STBIW_ASSERT(x) before the #include to avoid using assert.h. + You can #define STBIW_MALLOC(), STBIW_REALLOC(), and STBIW_FREE() to replace + malloc,realloc,free. + You can #define STBIW_MEMMOVE() to replace memmove() + You can #define STBIW_ZLIB_COMPRESS to use a custom zlib-style compress function + for PNG compression (instead of the builtin one), it must have the following signature: + unsigned char * my_compress(unsigned char *data, int data_len, int *out_len, int quality); + The returned data will be freed with STBIW_FREE() (free() by default), + so it must be heap allocated with STBIW_MALLOC() (malloc() by default), + +UNICODE: + + If compiling for Windows and you wish to use Unicode filenames, compile + with + #define STBIW_WINDOWS_UTF8 + and pass utf8-encoded filenames. Call stbiw_convert_wchar_to_utf8 to convert + Windows wchar_t filenames to utf8. + +USAGE: + + There are five functions, one for each image file format: + + int stbi_write_png(char const *filename, int w, int h, int comp, const void *data, int stride_in_bytes); + int stbi_write_bmp(char const *filename, int w, int h, int comp, const void *data); + int stbi_write_tga(char const *filename, int w, int h, int comp, const void *data); + int stbi_write_jpg(char const *filename, int w, int h, int comp, const void *data, int quality); + int stbi_write_hdr(char const *filename, int w, int h, int comp, const float *data); + + void stbi_flip_vertically_on_write(int flag); // flag is non-zero to flip data vertically + + There are also five equivalent functions that use an arbitrary write function. You are + expected to open/close your file-equivalent before and after calling these: + + int stbi_write_png_to_func(stbi_write_func *func, void *context, int w, int h, int comp, const void *data, int stride_in_bytes); + int stbi_write_bmp_to_func(stbi_write_func *func, void *context, int w, int h, int comp, const void *data); + int stbi_write_tga_to_func(stbi_write_func *func, void *context, int w, int h, int comp, const void *data); + int stbi_write_hdr_to_func(stbi_write_func *func, void *context, int w, int h, int comp, const float *data); + int stbi_write_jpg_to_func(stbi_write_func *func, void *context, int x, int y, int comp, const void *data, int quality); + + where the callback is: + void stbi_write_func(void *context, void *data, int size); + + You can configure it with these global variables: + int stbi_write_tga_with_rle; // defaults to true; set to 0 to disable RLE + int stbi_write_png_compression_level;// defaults to 8; set to higher for more compression + int stbi_write_force_png_filter; // defaults to -1; set to 0..5 to force a filter mode + + + You can define STBI_WRITE_NO_STDIO to disable the file variant of these + functions, so the library will not use stdio.h at all. However, this will + also disable HDR writing, because it requires stdio for formatted output. + + Each function returns 0 on failure and non-0 on success. + + The functions create an image file defined by the parameters. The image + is a rectangle of pixels stored from left-to-right, top-to-bottom. + Each pixel contains 'comp' channels of data stored interleaved with 8-bits + per channel, in the following order: 1=Y, 2=YA, 3=RGB, 4=RGBA. (Y is + monochrome color.) The rectangle is 'w' pixels wide and 'h' pixels tall. + The *data pointer points to the first byte of the top-left-most pixel. + For PNG, "stride_in_bytes" is the distance in bytes from the first byte of + a row of pixels to the first byte of the next row of pixels. + + PNG creates output files with the same number of components as the input. + The BMP format expands Y to RGB in the file format and does not + output alpha. + + PNG supports writing rectangles of data even when the bytes storing rows of + data are not consecutive in memory (e.g. sub-rectangles of a larger image), + by supplying the stride between the beginning of adjacent rows. The other + formats do not. (Thus you cannot write a native-format BMP through the BMP + writer, both because it is in BGR order and because it may have padding + at the end of the line.) + + PNG allows you to set the deflate compression level by setting the global + variable 'stbi_write_png_compression_level' (it defaults to 8). + + HDR expects linear float data. Since the format is always 32-bit rgb(e) + data, alpha (if provided) is discarded, and for monochrome data it is + replicated across all three channels. + + TGA supports RLE or non-RLE compressed data. To use non-RLE-compressed + data, set the global variable 'stbi_write_tga_with_rle' to 0. + + JPEG does ignore alpha channels in input data; quality is between 1 and 100. + Higher quality looks better but results in a bigger image. + JPEG baseline (no JPEG progressive). + +CREDITS: + + + Sean Barrett -PNG/BMP/TGA + Baldur Karlsson-HDR + Jean-Sebastien Guay-TGA monochrome + Tim Kelsey -misc enhancements + Alan Hickman -TGA RLE + Emmanuel Julien-initial file IO callback implementation + Jon Olick -original jo_jpeg.cpp code + Daniel Gibson -integrate JPEG, allow external zlib + Aarni Koskela -allow choosing PNG filter + + bugfixes: + github:Chribba + Guillaume Chereau + github:jry2 + github:romigrou + Sergio Gonzalez + Jonas Karlsson + Filip Wasil + Thatcher Ulrich + github:poppolopoppo + Patrick Boettcher + github:xeekworx + Cap Petschulat + Simon Rodriguez + Ivan Tikhonov + github:ignotion + Adam Schackart + +LICENSE + + See end of file for license information. + +*/ + +#ifndef INCLUDE_STB_IMAGE_WRITE_H +#define INCLUDE_STB_IMAGE_WRITE_H + +#include + +// if STB_IMAGE_WRITE_STATIC causes problems, try defining STBIWDEF to 'inline' or 'static inline' +#ifndef STBIWDEF +#ifdef STB_IMAGE_WRITE_STATIC +#define STBIWDEF static +#else +#ifdef __cplusplus +#define STBIWDEF extern "C" +#else +#define STBIWDEF extern +#endif +#endif +#endif + +#ifndef STB_IMAGE_WRITE_STATIC // C++ forbids static forward declarations +extern int stbi_write_tga_with_rle; +extern int stbi_write_png_compression_level; +extern int stbi_write_force_png_filter; +#endif + +#ifndef STBI_WRITE_NO_STDIO +STBIWDEF int stbi_write_png(char const *filename, int w, int h, int comp, const void *data, int stride_in_bytes); +STBIWDEF int stbi_write_bmp(char const *filename, int w, int h, int comp, const void *data); +STBIWDEF int stbi_write_tga(char const *filename, int w, int h, int comp, const void *data); +STBIWDEF int stbi_write_hdr(char const *filename, int w, int h, int comp, const float *data); +STBIWDEF int stbi_write_jpg(char const *filename, int x, int y, int comp, const void *data, int quality); + +#ifdef STBI_WINDOWS_UTF8 +STBIWDEF int stbiw_convert_wchar_to_utf8(char *buffer, size_t bufferlen, const wchar_t* input); +#endif +#endif + +typedef void stbi_write_func(void *context, void *data, int size); + +STBIWDEF int stbi_write_png_to_func(stbi_write_func *func, void *context, int w, int h, int comp, const void *data, int stride_in_bytes); +STBIWDEF int stbi_write_bmp_to_func(stbi_write_func *func, void *context, int w, int h, int comp, const void *data); +STBIWDEF int stbi_write_tga_to_func(stbi_write_func *func, void *context, int w, int h, int comp, const void *data); +STBIWDEF int stbi_write_hdr_to_func(stbi_write_func *func, void *context, int w, int h, int comp, const float *data); +STBIWDEF int stbi_write_jpg_to_func(stbi_write_func *func, void *context, int x, int y, int comp, const void *data, int quality); + +STBIWDEF void stbi_flip_vertically_on_write(int flip_boolean); + +#endif//INCLUDE_STB_IMAGE_WRITE_H + +#ifdef STB_IMAGE_WRITE_IMPLEMENTATION + +#ifdef _WIN32 + #ifndef _CRT_SECURE_NO_WARNINGS + #define _CRT_SECURE_NO_WARNINGS + #endif + #ifndef _CRT_NONSTDC_NO_DEPRECATE + #define _CRT_NONSTDC_NO_DEPRECATE + #endif +#endif + +#ifndef STBI_WRITE_NO_STDIO +#include +#endif // STBI_WRITE_NO_STDIO + +#include +#include +#include +#include + +#if defined(STBIW_MALLOC) && defined(STBIW_FREE) && (defined(STBIW_REALLOC) || defined(STBIW_REALLOC_SIZED)) +// ok +#elif !defined(STBIW_MALLOC) && !defined(STBIW_FREE) && !defined(STBIW_REALLOC) && !defined(STBIW_REALLOC_SIZED) +// ok +#else +#error "Must define all or none of STBIW_MALLOC, STBIW_FREE, and STBIW_REALLOC (or STBIW_REALLOC_SIZED)." +#endif + +#ifndef STBIW_MALLOC +#define STBIW_MALLOC(sz)malloc(sz) +#define STBIW_REALLOC(p,newsz) realloc(p,newsz) +#define STBIW_FREE(p) free(p) +#endif + +#ifndef STBIW_REALLOC_SIZED +#define STBIW_REALLOC_SIZED(p,oldsz,newsz) STBIW_REALLOC(p,newsz) +#endif + + +#ifndef STBIW_MEMMOVE +#define STBIW_MEMMOVE(a,b,sz) memmove(a,b,sz) +#endif + + +#ifndef STBIW_ASSERT +#include +#define STBIW_ASSERT(x) assert(x) +#endif + +#define STBIW_UCHAR(x) (unsigned char) ((x) & 0xff) + +#ifdef STB_IMAGE_WRITE_STATIC +static int stbi_write_png_compression_level = 8; +static int stbi_write_tga_with_rle = 1; +static int stbi_write_force_png_filter = -1; +#else +int stbi_write_png_compression_level = 8; +int stbi_write_tga_with_rle = 1; +int stbi_write_force_png_filter = -1; +#endif + +static int stbi__flip_vertically_on_write = 0; + +STBIWDEF void stbi_flip_vertically_on_write(int flag) +{ + stbi__flip_vertically_on_write = flag; +} + +typedef struct +{ + stbi_write_func *func; + void *context; +} stbi__write_context; + +// initialize a callback-based context +static void stbi__start_write_callbacks(stbi__write_context *s, stbi_write_func *c, void *context) +{ + s->func= c; + s->context = context; +} + +#ifndef STBI_WRITE_NO_STDIO + +static void stbi__stdio_write(void *context, void *data, int size) +{ + fwrite(data,1,size,(FILE*) context); +} + +#if defined(_MSC_VER) && defined(STBI_WINDOWS_UTF8) +#ifdef __cplusplus +#define STBIW_EXTERN extern "C" +#else +#define STBIW_EXTERN extern +#endif +STBIW_EXTERN __declspec(dllimport) int __stdcall MultiByteToWideChar(unsigned int cp, unsigned long flags, const char *str, int cbmb, wchar_t *widestr, int cchwide); +STBIW_EXTERN __declspec(dllimport) int __stdcall WideCharToMultiByte(unsigned int cp, unsigned long flags, const wchar_t *widestr, int cchwide, char *str, int cbmb, const char *defchar, int *used_default); + +STBIWDEF int stbiw_convert_wchar_to_utf8(char *buffer, size_t bufferlen, const wchar_t* input) +{ + return WideCharToMultiByte(65001 /* UTF8 */, 0, input, -1, buffer, (int) bufferlen, NULL, NULL); +} +#endif + +static FILE *stbiw__fopen(char const *filename, char const *mode) +{ + FILE *f; +#if defined(_MSC_VER) && defined(STBI_WINDOWS_UTF8) + wchar_t wMode[64]; + wchar_t wFilename[1024]; + if (0 == MultiByteToWideChar(65001 /* UTF8 */, 0, filename, -1, wFilename, sizeof(wFilename))) + return 0; + + if (0 == MultiByteToWideChar(65001 /* UTF8 */, 0, mode, -1, wMode, sizeof(wMode))) + return 0; + +#if _MSC_VER >= 1400 + if (0 != _wfopen_s(&f, wFilename, wMode)) + f = 0; +#else + f = _wfopen(wFilename, wMode); +#endif + +#elif defined(_MSC_VER) && _MSC_VER >= 1400 + if (0 != fopen_s(&f, filename, mode)) + f=0; +#else + f = fopen(filename, mode); +#endif + return f; +} + +static int stbi__start_write_file(stbi__write_context *s, const char *filename) +{ + FILE *f = stbiw__fopen(filename, "wb"); + stbi__start_write_callbacks(s, stbi__stdio_write, (void *) f); + return f != NULL; +} + +static void stbi__end_write_file(stbi__write_context *s) +{ + fclose((FILE *)s->context); +} + +#endif // !STBI_WRITE_NO_STDIO + +typedef unsigned int stbiw_uint32; +typedef int stb_image_write_test[sizeof(stbiw_uint32)==4 ? 1 : -1]; + +static void stbiw__writefv(stbi__write_context *s, const char *fmt, va_list v) +{ + while (*fmt) { + switch (*fmt++) { + case ' ': break; + case '1': { unsigned char x = STBIW_UCHAR(va_arg(v, int)); + s->func(s->context,&x,1); + break; } + case '2': { int x = va_arg(v,int); + unsigned char b[2]; + b[0] = STBIW_UCHAR(x); + b[1] = STBIW_UCHAR(x>>8); + s->func(s->context,b,2); + break; } + case '4': { stbiw_uint32 x = va_arg(v,int); + unsigned char b[4]; + b[0]=STBIW_UCHAR(x); + b[1]=STBIW_UCHAR(x>>8); + b[2]=STBIW_UCHAR(x>>16); + b[3]=STBIW_UCHAR(x>>24); + s->func(s->context,b,4); + break; } + default: +STBIW_ASSERT(0); +return; + } + } +} + +static void stbiw__writef(stbi__write_context *s, const char *fmt, ...) +{ + va_list v; + va_start(v, fmt); + stbiw__writefv(s, fmt, v); + va_end(v); +} + +static void stbiw__putc(stbi__write_context *s, unsigned char c) +{ + s->func(s->context, &c, 1); +} + +static void stbiw__write3(stbi__write_context *s, unsigned char a, unsigned char b, unsigned char c) +{ + unsigned char arr[3]; + arr[0] = a; arr[1] = b; arr[2] = c; + s->func(s->context, arr, 3); +} + +static void stbiw__write_pixel(stbi__write_context *s, int rgb_dir, int comp, int write_alpha, int expand_mono, unsigned char *d) +{ + unsigned char bg[3] = { 255, 0, 255}, px[3]; + int k; + + if (write_alpha < 0) + s->func(s->context, &d[comp - 1], 1); + + switch (comp) { + case 2: // 2 pixels = mono + alpha, alpha is written separately, so same as 1-channel case + case 1: + if (expand_mono) +stbiw__write3(s, d[0], d[0], d[0]); // monochrome bmp + else +s->func(s->context, d, 1); // monochrome TGA + break; + case 4: + if (!write_alpha) { +// composite against pink background +for (k = 0; k < 3; ++k) + px[k] = bg[k] + ((d[k] - bg[k]) * d[3]) / 255; +stbiw__write3(s, px[1 - rgb_dir], px[1], px[1 + rgb_dir]); +break; + } + /* FALLTHROUGH */ + case 3: + stbiw__write3(s, d[1 - rgb_dir], d[1], d[1 + rgb_dir]); + break; + } + if (write_alpha > 0) + s->func(s->context, &d[comp - 1], 1); +} + +static void stbiw__write_pixels(stbi__write_context *s, int rgb_dir, int vdir, int x, int y, int comp, void *data, int write_alpha, int scanline_pad, int expand_mono) +{ + stbiw_uint32 zero = 0; + int i,j, j_end; + + if (y <= 0) + return; + + if (stbi__flip_vertically_on_write) + vdir *= -1; + + if (vdir < 0) { + j_end = -1; j = y-1; + } else { + j_end = y; j = 0; + } + + for (; j != j_end; j += vdir) { + for (i=0; i < x; ++i) { + unsigned char *d = (unsigned char *) data + (j*x+i)*comp; + stbiw__write_pixel(s, rgb_dir, comp, write_alpha, expand_mono, d); + } + s->func(s->context, &zero, scanline_pad); + } +} + +static int stbiw__outfile(stbi__write_context *s, int rgb_dir, int vdir, int x, int y, int comp, int expand_mono, void *data, int alpha, int pad, const char *fmt, ...) +{ + if (y < 0 || x < 0) { + return 0; + } else { + va_list v; + va_start(v, fmt); + stbiw__writefv(s, fmt, v); + va_end(v); + stbiw__write_pixels(s,rgb_dir,vdir,x,y,comp,data,alpha,pad, expand_mono); + return 1; + } +} + +static int stbi_write_bmp_core(stbi__write_context *s, int x, int y, int comp, const void *data) +{ + int pad = (-x*3) & 3; + return stbiw__outfile(s,-1,-1,x,y,comp,1,(void *) data,0,pad, + "11 4 22 4" "4 44 22 444444", + 'B', 'M', 14+40+(x*3+pad)*y, 0,0, 14+40, // file header +40, x,y, 1,24, 0,0,0,0,0,0); // bitmap header +} + +STBIWDEF int stbi_write_bmp_to_func(stbi_write_func *func, void *context, int x, int y, int comp, const void *data) +{ + stbi__write_context s; + stbi__start_write_callbacks(&s, func, context); + return stbi_write_bmp_core(&s, x, y, comp, data); +} + +#ifndef STBI_WRITE_NO_STDIO +STBIWDEF int stbi_write_bmp(char const *filename, int x, int y, int comp, const void *data) +{ + stbi__write_context s; + if (stbi__start_write_file(&s,filename)) { + int r = stbi_write_bmp_core(&s, x, y, comp, data); + stbi__end_write_file(&s); + return r; + } else + return 0; +} +#endif //!STBI_WRITE_NO_STDIO + +static int stbi_write_tga_core(stbi__write_context *s, int x, int y, int comp, void *data) +{ + int has_alpha = (comp == 2 || comp == 4); + int colorbytes = has_alpha ? comp-1 : comp; + int format = colorbytes < 2 ? 3 : 2; // 3 color channels (RGB/RGBA) = 2, 1 color channel (Y/YA) = 3 + + if (y < 0 || x < 0) + return 0; + + if (!stbi_write_tga_with_rle) { + return stbiw__outfile(s, -1, -1, x, y, comp, 0, (void *) data, has_alpha, 0, + "111 221 2222 11", 0, 0, format, 0, 0, 0, 0, 0, x, y, (colorbytes + has_alpha) * 8, has_alpha * 8); + } else { + int i,j,k; + int jend, jdir; + + stbiw__writef(s, "111 221 2222 11", 0,0,format+8, 0,0,0, 0,0,x,y, (colorbytes + has_alpha) * 8, has_alpha * 8); + + if (stbi__flip_vertically_on_write) { + j = 0; + jend = y; + jdir = 1; + } else { + j = y-1; + jend = -1; + jdir = -1; + } + for (; j != jend; j += jdir) { + unsigned char *row = (unsigned char *) data + j * x * comp; + int len; + + for (i = 0; i < x; i += len) { +unsigned char *begin = row + i * comp; +int diff = 1; +len = 1; + +if (i < x - 1) { + ++len; + diff = memcmp(begin, row + (i + 1) * comp, comp); + if (diff) { + const unsigned char *prev = begin; + for (k = i + 2; k < x && len < 128; ++k) { + if (memcmp(prev, row + k * comp, comp)) { +prev += comp; +++len; + } else { +--len; +break; + } + } + } else { + for (k = i + 2; k < x && len < 128; ++k) { + if (!memcmp(begin, row + k * comp, comp)) { +++len; + } else { +break; + } + } + } +} + +if (diff) { + unsigned char header = STBIW_UCHAR(len - 1); + s->func(s->context, &header, 1); + for (k = 0; k < len; ++k) { + stbiw__write_pixel(s, -1, comp, has_alpha, 0, begin + k * comp); + } +} else { + unsigned char header = STBIW_UCHAR(len - 129); + s->func(s->context, &header, 1); + stbiw__write_pixel(s, -1, comp, has_alpha, 0, begin); +} + } + } + } + return 1; +} + +STBIWDEF int stbi_write_tga_to_func(stbi_write_func *func, void *context, int x, int y, int comp, const void *data) +{ + stbi__write_context s; + stbi__start_write_callbacks(&s, func, context); + return stbi_write_tga_core(&s, x, y, comp, (void *) data); +} + +#ifndef STBI_WRITE_NO_STDIO +STBIWDEF int stbi_write_tga(char const *filename, int x, int y, int comp, const void *data) +{ + stbi__write_context s; + if (stbi__start_write_file(&s,filename)) { + int r = stbi_write_tga_core(&s, x, y, comp, (void *) data); + stbi__end_write_file(&s); + return r; + } else + return 0; +} +#endif + +// ************************************************************************************************* +// Radiance RGBE HDR writer +// by Baldur Karlsson + +#define stbiw__max(a, b) ((a) > (b) ? (a) : (b)) + +static void stbiw__linear_to_rgbe(unsigned char *rgbe, float *linear) +{ + int exponent; + float maxcomp = stbiw__max(linear[0], stbiw__max(linear[1], linear[2])); + + if (maxcomp < 1e-32f) { + rgbe[0] = rgbe[1] = rgbe[2] = rgbe[3] = 0; + } else { + float normalize = (float) frexp(maxcomp, &exponent) * 256.0f/maxcomp; + + rgbe[0] = (unsigned char)(linear[0] * normalize); + rgbe[1] = (unsigned char)(linear[1] * normalize); + rgbe[2] = (unsigned char)(linear[2] * normalize); + rgbe[3] = (unsigned char)(exponent + 128); + } +} + +static void stbiw__write_run_data(stbi__write_context *s, int length, unsigned char databyte) +{ + unsigned char lengthbyte = STBIW_UCHAR(length+128); + STBIW_ASSERT(length+128 <= 255); + s->func(s->context, &lengthbyte, 1); + s->func(s->context, &databyte, 1); +} + +static void stbiw__write_dump_data(stbi__write_context *s, int length, unsigned char *data) +{ + unsigned char lengthbyte = STBIW_UCHAR(length); + STBIW_ASSERT(length <= 128); // inconsistent with spec but consistent with official code + s->func(s->context, &lengthbyte, 1); + s->func(s->context, data, length); +} + +static void stbiw__write_hdr_scanline(stbi__write_context *s, int width, int ncomp, unsigned char *scratch, float *scanline) +{ + unsigned char scanlineheader[4] = { 2, 2, 0, 0 }; + unsigned char rgbe[4]; + float linear[3]; + int x; + + scanlineheader[2] = (width&0xff00)>>8; + scanlineheader[3] = (width&0x00ff); + + /* skip RLE for images too small or large */ + if (width < 8 || width >= 32768) { + for (x=0; x < width; x++) { + switch (ncomp) { +case 4: /* fallthrough */ +case 3: linear[2] = scanline[x*ncomp + 2]; +linear[1] = scanline[x*ncomp + 1]; +linear[0] = scanline[x*ncomp + 0]; +break; +default: +linear[0] = linear[1] = linear[2] = scanline[x*ncomp + 0]; +break; + } + stbiw__linear_to_rgbe(rgbe, linear); + s->func(s->context, rgbe, 4); + } + } else { + int c,r; + /* encode into scratch buffer */ + for (x=0; x < width; x++) { + switch(ncomp) { +case 4: /* fallthrough */ +case 3: linear[2] = scanline[x*ncomp + 2]; +linear[1] = scanline[x*ncomp + 1]; +linear[0] = scanline[x*ncomp + 0]; +break; +default: +linear[0] = linear[1] = linear[2] = scanline[x*ncomp + 0]; +break; + } + stbiw__linear_to_rgbe(rgbe, linear); + scratch[x + width*0] = rgbe[0]; + scratch[x + width*1] = rgbe[1]; + scratch[x + width*2] = rgbe[2]; + scratch[x + width*3] = rgbe[3]; + } + + s->func(s->context, scanlineheader, 4); + + /* RLE each component separately */ + for (c=0; c < 4; c++) { + unsigned char *comp = &scratch[width*c]; + + x = 0; + while (x < width) { +// find first run +r = x; +while (r+2 < width) { + if (comp[r] == comp[r+1] && comp[r] == comp[r+2]) + break; + ++r; +} +if (r+2 >= width) + r = width; +// dump up to first run +while (x < r) { + int len = r-x; + if (len > 128) len = 128; + stbiw__write_dump_data(s, len, &comp[x]); + x += len; +} +// if there's a run, output it +if (r+2 < width) { // same test as what we break out of in search loop, so only true if we break'd + // find next byte after run + while (r < width && comp[r] == comp[x]) + ++r; + // output run up to r + while (x < r) { + int len = r-x; + if (len > 127) len = 127; + stbiw__write_run_data(s, len, comp[x]); + x += len; + } +} + } + } + } +} + +static int stbi_write_hdr_core(stbi__write_context *s, int x, int y, int comp, float *data) +{ + if (y <= 0 || x <= 0 || data == NULL) + return 0; + else { + // Each component is stored separately. Allocate scratch space for full output scanline. + unsigned char *scratch = (unsigned char *) STBIW_MALLOC(x*4); + int i, len; + char buffer[128]; + char header[] = "#?RADIANCE\n# Written by stb_image_write.h\nFORMAT=32-bit_rle_rgbe\n"; + s->func(s->context, header, sizeof(header)-1); + +#ifdef __STDC_WANT_SECURE_LIB__ + len = sprintf_s(buffer, sizeof(buffer), "EXPOSURE= 1.0000000000000\n\n-Y %d +X %d\n", y, x); +#else + len = sprintf(buffer, "EXPOSURE= 1.0000000000000\n\n-Y %d +X %d\n", y, x); +#endif + s->func(s->context, buffer, len); + + for(i=0; i < y; i++) + stbiw__write_hdr_scanline(s, x, comp, scratch, data + comp*x*(stbi__flip_vertically_on_write ? y-1-i : i)); + STBIW_FREE(scratch); + return 1; + } +} + +STBIWDEF int stbi_write_hdr_to_func(stbi_write_func *func, void *context, int x, int y, int comp, const float *data) +{ + stbi__write_context s; + stbi__start_write_callbacks(&s, func, context); + return stbi_write_hdr_core(&s, x, y, comp, (float *) data); +} + +#ifndef STBI_WRITE_NO_STDIO +STBIWDEF int stbi_write_hdr(char const *filename, int x, int y, int comp, const float *data) +{ + stbi__write_context s; + if (stbi__start_write_file(&s,filename)) { + int r = stbi_write_hdr_core(&s, x, y, comp, (float *) data); + stbi__end_write_file(&s); + return r; + } else + return 0; +} +#endif // STBI_WRITE_NO_STDIO + + +////////////////////////////////////////////////////////////////////////////// +// +// PNG writer +// + +#ifndef STBIW_ZLIB_COMPRESS +// stretchy buffer; stbiw__sbpush() == vector<>::push_back() -- stbiw__sbcount() == vector<>::size() +#define stbiw__sbraw(a) ((int *) (void *) (a) - 2) +#define stbiw__sbm(a) stbiw__sbraw(a)[0] +#define stbiw__sbn(a) stbiw__sbraw(a)[1] + +#define stbiw__sbneedgrow(a,n) ((a)==0 || stbiw__sbn(a)+n >= stbiw__sbm(a)) +#define stbiw__sbmaybegrow(a,n) (stbiw__sbneedgrow(a,(n)) ? stbiw__sbgrow(a,n) : 0) +#define stbiw__sbgrow(a,n) stbiw__sbgrowf((void **) &(a), (n), sizeof(*(a))) + +#define stbiw__sbpush(a, v) (stbiw__sbmaybegrow(a,1), (a)[stbiw__sbn(a)++] = (v)) +#define stbiw__sbcount(a)((a) ? stbiw__sbn(a) : 0) +#define stbiw__sbfree(a) ((a) ? STBIW_FREE(stbiw__sbraw(a)),0 : 0) + +static void *stbiw__sbgrowf(void **arr, int increment, int itemsize) +{ + int m = *arr ? 2*stbiw__sbm(*arr)+increment : increment+1; + void *p = STBIW_REALLOC_SIZED(*arr ? stbiw__sbraw(*arr) : 0, *arr ? (stbiw__sbm(*arr)*itemsize + sizeof(int)*2) : 0, itemsize * m + sizeof(int)*2); + STBIW_ASSERT(p); + if (p) { + if (!*arr) ((int *) p)[1] = 0; + *arr = (void *) ((int *) p + 2); + stbiw__sbm(*arr) = m; + } + return *arr; +} + +static unsigned char *stbiw__zlib_flushf(unsigned char *data, unsigned int *bitbuffer, int *bitcount) +{ + while (*bitcount >= 8) { + stbiw__sbpush(data, STBIW_UCHAR(*bitbuffer)); + *bitbuffer >>= 8; + *bitcount -= 8; + } + return data; +} + +static int stbiw__zlib_bitrev(int code, int codebits) +{ + int res=0; + while (codebits--) { + res = (res << 1) | (code & 1); + code >>= 1; + } + return res; +} + +static unsigned int stbiw__zlib_countm(unsigned char *a, unsigned char *b, int limit) +{ + int i; + for (i=0; i < limit && i < 258; ++i) + if (a[i] != b[i]) break; + return i; +} + +static unsigned int stbiw__zhash(unsigned char *data) +{ + stbiw_uint32 hash = data[0] + (data[1] << 8) + (data[2] << 16); + hash ^= hash << 3; + hash += hash >> 5; + hash ^= hash << 4; + hash += hash >> 17; + hash ^= hash << 25; + hash += hash >> 6; + return hash; +} + +#define stbiw__zlib_flush() (out = stbiw__zlib_flushf(out, &bitbuf, &bitcount)) +#define stbiw__zlib_add(code,codebits) \ + (bitbuf |= (code) << bitcount, bitcount += (codebits), stbiw__zlib_flush()) +#define stbiw__zlib_huffa(b,c) stbiw__zlib_add(stbiw__zlib_bitrev(b,c),c) +// default huffman tables +#define stbiw__zlib_huff1(n) stbiw__zlib_huffa(0x30 + (n), 8) +#define stbiw__zlib_huff2(n) stbiw__zlib_huffa(0x190 + (n)-144, 9) +#define stbiw__zlib_huff3(n) stbiw__zlib_huffa(0 + (n)-256,7) +#define stbiw__zlib_huff4(n) stbiw__zlib_huffa(0xc0 + (n)-280,8) +#define stbiw__zlib_huff(n) ((n) <= 143 ? stbiw__zlib_huff1(n) : (n) <= 255 ? stbiw__zlib_huff2(n) : (n) <= 279 ? stbiw__zlib_huff3(n) : stbiw__zlib_huff4(n)) +#define stbiw__zlib_huffb(n) ((n) <= 143 ? stbiw__zlib_huff1(n) : stbiw__zlib_huff2(n)) + +#define stbiw__ZHASH 16384 + +#endif // STBIW_ZLIB_COMPRESS + +STBIWDEF unsigned char * stbi_zlib_compress(unsigned char *data, int data_len, int *out_len, int quality) +{ +#ifdef STBIW_ZLIB_COMPRESS + // user provided a zlib compress implementation, use that + return STBIW_ZLIB_COMPRESS(data, data_len, out_len, quality); +#else // use builtin + static unsigned short lengthc[] = { 3,4,5,6,7,8,9,10,11,13,15,17,19,23,27,31,35,43,51,59,67,83,99,115,131,163,195,227,258, 259 }; + static unsigned char lengtheb[]= { 0,0,0,0,0,0,0, 0, 1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3, 4, 4, 4, 4, 5, 5, 5, 5, 0 }; + static unsigned short distc[] = { 1,2,3,4,5,7,9,13,17,25,33,49,65,97,129,193,257,385,513,769,1025,1537,2049,3073,4097,6145,8193,12289,16385,24577, 32768 }; + static unsigned char disteb[] = { 0,0,0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7,8,8,9,9,10,10,11,11,12,12,13,13 }; + unsigned int bitbuf=0; + int i,j, bitcount=0; + unsigned char *out = NULL; + unsigned char ***hash_table = (unsigned char***) STBIW_MALLOC(stbiw__ZHASH * sizeof(unsigned char**)); + if (hash_table == NULL) + return NULL; + if (quality < 5) quality = 5; + + stbiw__sbpush(out, 0x78); // DEFLATE 32K window + stbiw__sbpush(out, 0x5e); // FLEVEL = 1 + stbiw__zlib_add(1,1); // BFINAL = 1 + stbiw__zlib_add(1,2); // BTYPE = 1 -- fixed huffman + + for (i=0; i < stbiw__ZHASH; ++i) + hash_table[i] = NULL; + + i=0; + while (i < data_len-3) { + // hash next 3 bytes of data to be compressed + int h = stbiw__zhash(data+i)&(stbiw__ZHASH-1), best=3; + unsigned char *bestloc = 0; + unsigned char **hlist = hash_table[h]; + int n = stbiw__sbcount(hlist); + for (j=0; j < n; ++j) { + if (hlist[j]-data > i-32768) { // if entry lies within window +int d = stbiw__zlib_countm(hlist[j], data+i, data_len-i); +if (d >= best) { best=d; bestloc=hlist[j]; } + } + } + // when hash table entry is too long, delete half the entries + if (hash_table[h] && stbiw__sbn(hash_table[h]) == 2*quality) { + STBIW_MEMMOVE(hash_table[h], hash_table[h]+quality, sizeof(hash_table[h][0])*quality); + stbiw__sbn(hash_table[h]) = quality; + } + stbiw__sbpush(hash_table[h],data+i); + + if (bestloc) { + // "lazy matching" - check match at *next* byte, and if it's better, do cur byte as literal + h = stbiw__zhash(data+i+1)&(stbiw__ZHASH-1); + hlist = hash_table[h]; + n = stbiw__sbcount(hlist); + for (j=0; j < n; ++j) { +if (hlist[j]-data > i-32767) { + int e = stbiw__zlib_countm(hlist[j], data+i+1, data_len-i-1); + if (e > best) { // if next match is better, bail on current match + bestloc = NULL; + break; + } +} + } + } + + if (bestloc) { + int d = (int) (data+i - bestloc); // distance back + STBIW_ASSERT(d <= 32767 && best <= 258); + for (j=0; best > lengthc[j+1]-1; ++j); + stbiw__zlib_huff(j+257); + if (lengtheb[j]) stbiw__zlib_add(best - lengthc[j], lengtheb[j]); + for (j=0; d > distc[j+1]-1; ++j); + stbiw__zlib_add(stbiw__zlib_bitrev(j,5),5); + if (disteb[j]) stbiw__zlib_add(d - distc[j], disteb[j]); + i += best; + } else { + stbiw__zlib_huffb(data[i]); + ++i; + } + } + // write out final bytes + for (;i < data_len; ++i) + stbiw__zlib_huffb(data[i]); + stbiw__zlib_huff(256); // end of block + // pad with 0 bits to byte boundary + while (bitcount) + stbiw__zlib_add(0,1); + + for (i=0; i < stbiw__ZHASH; ++i) + (void) stbiw__sbfree(hash_table[i]); + STBIW_FREE(hash_table); + + { + // compute adler32 on input + unsigned int s1=1, s2=0; + int blocklen = (int) (data_len % 5552); + j=0; + while (j < data_len) { + for (i=0; i < blocklen; ++i) { s1 += data[j+i]; s2 += s1; } + s1 %= 65521; s2 %= 65521; + j += blocklen; + blocklen = 5552; + } + stbiw__sbpush(out, STBIW_UCHAR(s2 >> 8)); + stbiw__sbpush(out, STBIW_UCHAR(s2)); + stbiw__sbpush(out, STBIW_UCHAR(s1 >> 8)); + stbiw__sbpush(out, STBIW_UCHAR(s1)); + } + *out_len = stbiw__sbn(out); + // make returned pointer freeable + STBIW_MEMMOVE(stbiw__sbraw(out), out, *out_len); + return (unsigned char *) stbiw__sbraw(out); +#endif // STBIW_ZLIB_COMPRESS +} + +static unsigned int stbiw__crc32(unsigned char *buffer, int len) +{ +#ifdef STBIW_CRC32 +return STBIW_CRC32(buffer, len); +#else + static unsigned int crc_table[256] = + { + 0x00000000, 0x77073096, 0xEE0E612C, 0x990951BA, 0x076DC419, 0x706AF48F, 0xE963A535, 0x9E6495A3, + 0x0eDB8832, 0x79DCB8A4, 0xE0D5E91E, 0x97D2D988, 0x09B64C2B, 0x7EB17CBD, 0xE7B82D07, 0x90BF1D91, + 0x1DB71064, 0x6AB020F2, 0xF3B97148, 0x84BE41DE, 0x1ADAD47D, 0x6DDDE4EB, 0xF4D4B551, 0x83D385C7, + 0x136C9856, 0x646BA8C0, 0xFD62F97A, 0x8A65C9EC, 0x14015C4F, 0x63066CD9, 0xFA0F3D63, 0x8D080DF5, + 0x3B6E20C8, 0x4C69105E, 0xD56041E4, 0xA2677172, 0x3C03E4D1, 0x4B04D447, 0xD20D85FD, 0xA50AB56B, + 0x35B5A8FA, 0x42B2986C, 0xDBBBC9D6, 0xACBCF940, 0x32D86CE3, 0x45DF5C75, 0xDCD60DCF, 0xABD13D59, + 0x26D930AC, 0x51DE003A, 0xC8D75180, 0xBFD06116, 0x21B4F4B5, 0x56B3C423, 0xCFBA9599, 0xB8BDA50F, + 0x2802B89E, 0x5F058808, 0xC60CD9B2, 0xB10BE924, 0x2F6F7C87, 0x58684C11, 0xC1611DAB, 0xB6662D3D, + 0x76DC4190, 0x01DB7106, 0x98D220BC, 0xEFD5102A, 0x71B18589, 0x06B6B51F, 0x9FBFE4A5, 0xE8B8D433, + 0x7807C9A2, 0x0F00F934, 0x9609A88E, 0xE10E9818, 0x7F6A0DBB, 0x086D3D2D, 0x91646C97, 0xE6635C01, + 0x6B6B51F4, 0x1C6C6162, 0x856530D8, 0xF262004E, 0x6C0695ED, 0x1B01A57B, 0x8208F4C1, 0xF50FC457, + 0x65B0D9C6, 0x12B7E950, 0x8BBEB8EA, 0xFCB9887C, 0x62DD1DDF, 0x15DA2D49, 0x8CD37CF3, 0xFBD44C65, + 0x4DB26158, 0x3AB551CE, 0xA3BC0074, 0xD4BB30E2, 0x4ADFA541, 0x3DD895D7, 0xA4D1C46D, 0xD3D6F4FB, + 0x4369E96A, 0x346ED9FC, 0xAD678846, 0xDA60B8D0, 0x44042D73, 0x33031DE5, 0xAA0A4C5F, 0xDD0D7CC9, + 0x5005713C, 0x270241AA, 0xBE0B1010, 0xC90C2086, 0x5768B525, 0x206F85B3, 0xB966D409, 0xCE61E49F, + 0x5EDEF90E, 0x29D9C998, 0xB0D09822, 0xC7D7A8B4, 0x59B33D17, 0x2EB40D81, 0xB7BD5C3B, 0xC0BA6CAD, + 0xEDB88320, 0x9ABFB3B6, 0x03B6E20C, 0x74B1D29A, 0xEAD54739, 0x9DD277AF, 0x04DB2615, 0x73DC1683, + 0xE3630B12, 0x94643B84, 0x0D6D6A3E, 0x7A6A5AA8, 0xE40ECF0B, 0x9309FF9D, 0x0A00AE27, 0x7D079EB1, + 0xF00F9344, 0x8708A3D2, 0x1E01F268, 0x6906C2FE, 0xF762575D, 0x806567CB, 0x196C3671, 0x6E6B06E7, + 0xFED41B76, 0x89D32BE0, 0x10DA7A5A, 0x67DD4ACC, 0xF9B9DF6F, 0x8EBEEFF9, 0x17B7BE43, 0x60B08ED5, + 0xD6D6A3E8, 0xA1D1937E, 0x38D8C2C4, 0x4FDFF252, 0xD1BB67F1, 0xA6BC5767, 0x3FB506DD, 0x48B2364B, + 0xD80D2BDA, 0xAF0A1B4C, 0x36034AF6, 0x41047A60, 0xDF60EFC3, 0xA867DF55, 0x316E8EEF, 0x4669BE79, + 0xCB61B38C, 0xBC66831A, 0x256FD2A0, 0x5268E236, 0xCC0C7795, 0xBB0B4703, 0x220216B9, 0x5505262F, + 0xC5BA3BBE, 0xB2BD0B28, 0x2BB45A92, 0x5CB36A04, 0xC2D7FFA7, 0xB5D0CF31, 0x2CD99E8B, 0x5BDEAE1D, + 0x9B64C2B0, 0xEC63F226, 0x756AA39C, 0x026D930A, 0x9C0906A9, 0xEB0E363F, 0x72076785, 0x05005713, + 0x95BF4A82, 0xE2B87A14, 0x7BB12BAE, 0x0CB61B38, 0x92D28E9B, 0xE5D5BE0D, 0x7CDCEFB7, 0x0BDBDF21, + 0x86D3D2D4, 0xF1D4E242, 0x68DDB3F8, 0x1FDA836E, 0x81BE16CD, 0xF6B9265B, 0x6FB077E1, 0x18B74777, + 0x88085AE6, 0xFF0F6A70, 0x66063BCA, 0x11010B5C, 0x8F659EFF, 0xF862AE69, 0x616BFFD3, 0x166CCF45, + 0xA00AE278, 0xD70DD2EE, 0x4E048354, 0x3903B3C2, 0xA7672661, 0xD06016F7, 0x4969474D, 0x3E6E77DB, + 0xAED16A4A, 0xD9D65ADC, 0x40DF0B66, 0x37D83BF0, 0xA9BCAE53, 0xDEBB9EC5, 0x47B2CF7F, 0x30B5FFE9, + 0xBDBDF21C, 0xCABAC28A, 0x53B39330, 0x24B4A3A6, 0xBAD03605, 0xCDD70693, 0x54DE5729, 0x23D967BF, + 0xB3667A2E, 0xC4614AB8, 0x5D681B02, 0x2A6F2B94, 0xB40BBE37, 0xC30C8EA1, 0x5A05DF1B, 0x2D02EF8D + }; + + unsigned int crc = ~0u; + int i; + for (i=0; i < len; ++i) + crc = (crc >> 8) ^ crc_table[buffer[i] ^ (crc & 0xff)]; + return ~crc; +#endif +} + +#define stbiw__wpng4(o,a,b,c,d) ((o)[0]=STBIW_UCHAR(a),(o)[1]=STBIW_UCHAR(b),(o)[2]=STBIW_UCHAR(c),(o)[3]=STBIW_UCHAR(d),(o)+=4) +#define stbiw__wp32(data,v) stbiw__wpng4(data, (v)>>24,(v)>>16,(v)>>8,(v)); +#define stbiw__wptag(data,s) stbiw__wpng4(data, s[0],s[1],s[2],s[3]) + +static void stbiw__wpcrc(unsigned char **data, int len) +{ + unsigned int crc = stbiw__crc32(*data - len - 4, len+4); + stbiw__wp32(*data, crc); +} + +static unsigned char stbiw__paeth(int a, int b, int c) +{ + int p = a + b - c, pa = abs(p-a), pb = abs(p-b), pc = abs(p-c); + if (pa <= pb && pa <= pc) return STBIW_UCHAR(a); + if (pb <= pc) return STBIW_UCHAR(b); + return STBIW_UCHAR(c); +} + +// @OPTIMIZE: provide an option that always forces left-predict or paeth predict +static void stbiw__encode_png_line(unsigned char *pixels, int stride_bytes, int width, int height, int y, int n, int filter_type, signed char *line_buffer) +{ + static int mapping[] = { 0,1,2,3,4 }; + static int firstmap[] = { 0,1,0,5,6 }; + int *mymap = (y != 0) ? mapping : firstmap; + int i; + int type = mymap[filter_type]; + unsigned char *z = pixels + stride_bytes * (stbi__flip_vertically_on_write ? height-1-y : y); + int signed_stride = stbi__flip_vertically_on_write ? -stride_bytes : stride_bytes; + + if (type==0) { + memcpy(line_buffer, z, width*n); + return; + } + + // first loop isn't optimized since it's just one pixel + for (i = 0; i < n; ++i) { + switch (type) { + case 1: line_buffer[i] = z[i]; break; + case 2: line_buffer[i] = z[i] - z[i-signed_stride]; break; + case 3: line_buffer[i] = z[i] - (z[i-signed_stride]>>1); break; + case 4: line_buffer[i] = (signed char) (z[i] - stbiw__paeth(0,z[i-signed_stride],0)); break; + case 5: line_buffer[i] = z[i]; break; + case 6: line_buffer[i] = z[i]; break; + } + } + switch (type) { + case 1: for (i=n; i < width*n; ++i) line_buffer[i] = z[i] - z[i-n]; break; + case 2: for (i=n; i < width*n; ++i) line_buffer[i] = z[i] - z[i-signed_stride]; break; + case 3: for (i=n; i < width*n; ++i) line_buffer[i] = z[i] - ((z[i-n] + z[i-signed_stride])>>1); break; + case 4: for (i=n; i < width*n; ++i) line_buffer[i] = z[i] - stbiw__paeth(z[i-n], z[i-signed_stride], z[i-signed_stride-n]); break; + case 5: for (i=n; i < width*n; ++i) line_buffer[i] = z[i] - (z[i-n]>>1); break; + case 6: for (i=n; i < width*n; ++i) line_buffer[i] = z[i] - stbiw__paeth(z[i-n], 0,0); break; + } +} + +STBIWDEF unsigned char *stbi_write_png_to_mem(const unsigned char *pixels, int stride_bytes, int x, int y, int n, int *out_len) +{ + int force_filter = stbi_write_force_png_filter; + int ctype[5] = { -1, 0, 4, 2, 6 }; + unsigned char sig[8] = { 137,80,78,71,13,10,26,10 }; + unsigned char *out,*o, *filt, *zlib; + signed char *line_buffer; + int j,zlen; + + if (stride_bytes == 0) + stride_bytes = x * n; + + if (force_filter >= 5) { + force_filter = -1; + } + + filt = (unsigned char *) STBIW_MALLOC((x*n+1) * y); if (!filt) return 0; + line_buffer = (signed char *) STBIW_MALLOC(x * n); if (!line_buffer) { STBIW_FREE(filt); return 0; } + for (j=0; j < y; ++j) { + int filter_type; + if (force_filter > -1) { + filter_type = force_filter; + stbiw__encode_png_line((unsigned char*)(pixels), stride_bytes, x, y, j, n, force_filter, line_buffer); + } else { // Estimate the best filter by running through all of them: + int best_filter = 0, best_filter_val = 0x7fffffff, est, i; + for (filter_type = 0; filter_type < 5; filter_type++) { +stbiw__encode_png_line((unsigned char*)(pixels), stride_bytes, x, y, j, n, filter_type, line_buffer); + +// Estimate the entropy of the line using this filter; the less, the better. +est = 0; +for (i = 0; i < x*n; ++i) { + est += abs((signed char) line_buffer[i]); +} +if (est < best_filter_val) { + best_filter_val = est; + best_filter = filter_type; +} + } + if (filter_type != best_filter) { // If the last iteration already got us the best filter, don't redo it +stbiw__encode_png_line((unsigned char*)(pixels), stride_bytes, x, y, j, n, best_filter, line_buffer); +filter_type = best_filter; + } + } + // when we get here, filter_type contains the filter type, and line_buffer contains the data + filt[j*(x*n+1)] = (unsigned char) filter_type; + STBIW_MEMMOVE(filt+j*(x*n+1)+1, line_buffer, x*n); + } + STBIW_FREE(line_buffer); + zlib = stbi_zlib_compress(filt, y*( x*n+1), &zlen, stbi_write_png_compression_level); + STBIW_FREE(filt); + if (!zlib) return 0; + + // each tag requires 12 bytes of overhead + out = (unsigned char *) STBIW_MALLOC(8 + 12+13 + 12+zlen + 12); + if (!out) return 0; + *out_len = 8 + 12+13 + 12+zlen + 12; + + o=out; + STBIW_MEMMOVE(o,sig,8); o+= 8; + stbiw__wp32(o, 13); // header length + stbiw__wptag(o, "IHDR"); + stbiw__wp32(o, x); + stbiw__wp32(o, y); + *o++ = 8; + *o++ = STBIW_UCHAR(ctype[n]); + *o++ = 0; + *o++ = 0; + *o++ = 0; + stbiw__wpcrc(&o,13); + + stbiw__wp32(o, zlen); + stbiw__wptag(o, "IDAT"); + STBIW_MEMMOVE(o, zlib, zlen); + o += zlen; + STBIW_FREE(zlib); + stbiw__wpcrc(&o, zlen); + + stbiw__wp32(o,0); + stbiw__wptag(o, "IEND"); + stbiw__wpcrc(&o,0); + + STBIW_ASSERT(o == out + *out_len); + + return out; +} + +#ifndef STBI_WRITE_NO_STDIO +STBIWDEF int stbi_write_png(char const *filename, int x, int y, int comp, const void *data, int stride_bytes) +{ + FILE *f; + int len; + unsigned char *png = stbi_write_png_to_mem((const unsigned char *) data, stride_bytes, x, y, comp, &len); + if (png == NULL) return 0; + + f = stbiw__fopen(filename, "wb"); + if (!f) { STBIW_FREE(png); return 0; } + fwrite(png, 1, len, f); + fclose(f); + STBIW_FREE(png); + return 1; +} +#endif + +STBIWDEF int stbi_write_png_to_func(stbi_write_func *func, void *context, int x, int y, int comp, const void *data, int stride_bytes) +{ + int len; + unsigned char *png = stbi_write_png_to_mem((const unsigned char *) data, stride_bytes, x, y, comp, &len); + if (png == NULL) return 0; + func(context, png, len); + STBIW_FREE(png); + return 1; +} + + +/* *************************************************************************** + * + * JPEG writer + * + * This is based on Jon Olick's jo_jpeg.cpp: + * public domain Simple, Minimalistic JPEG writer - http://www.jonolick.com/code.html + */ + +static const unsigned char stbiw__jpg_ZigZag[] = { 0,1,5,6,14,15,27,28,2,4,7,13,16,26,29,42,3,8,12,17,25,30,41,43,9,11,18, + 24,31,40,44,53,10,19,23,32,39,45,52,54,20,22,33,38,46,51,55,60,21,34,37,47,50,56,59,61,35,36,48,49,57,58,62,63 }; + +static void stbiw__jpg_writeBits(stbi__write_context *s, int *bitBufP, int *bitCntP, const unsigned short *bs) { + int bitBuf = *bitBufP, bitCnt = *bitCntP; + bitCnt += bs[1]; + bitBuf |= bs[0] << (24 - bitCnt); + while(bitCnt >= 8) { + unsigned char c = (bitBuf >> 16) & 255; + stbiw__putc(s, c); + if(c == 255) { + stbiw__putc(s, 0); + } + bitBuf <<= 8; + bitCnt -= 8; + } + *bitBufP = bitBuf; + *bitCntP = bitCnt; +} + +static void stbiw__jpg_DCT(float *d0p, float *d1p, float *d2p, float *d3p, float *d4p, float *d5p, float *d6p, float *d7p) { + float d0 = *d0p, d1 = *d1p, d2 = *d2p, d3 = *d3p, d4 = *d4p, d5 = *d5p, d6 = *d6p, d7 = *d7p; + float z1, z2, z3, z4, z5, z11, z13; + + float tmp0 = d0 + d7; + float tmp7 = d0 - d7; + float tmp1 = d1 + d6; + float tmp6 = d1 - d6; + float tmp2 = d2 + d5; + float tmp5 = d2 - d5; + float tmp3 = d3 + d4; + float tmp4 = d3 - d4; + + // Even part + float tmp10 = tmp0 + tmp3; // phase 2 + float tmp13 = tmp0 - tmp3; + float tmp11 = tmp1 + tmp2; + float tmp12 = tmp1 - tmp2; + + d0 = tmp10 + tmp11; // phase 3 + d4 = tmp10 - tmp11; + + z1 = (tmp12 + tmp13) * 0.707106781f; // c4 + d2 = tmp13 + z1; // phase 5 + d6 = tmp13 - z1; + + // Odd part + tmp10 = tmp4 + tmp5; // phase 2 + tmp11 = tmp5 + tmp6; + tmp12 = tmp6 + tmp7; + + // The rotator is modified from fig 4-8 to avoid extra negations. + z5 = (tmp10 - tmp12) * 0.382683433f; // c6 + z2 = tmp10 * 0.541196100f + z5; // c2-c6 + z4 = tmp12 * 1.306562965f + z5; // c2+c6 + z3 = tmp11 * 0.707106781f; // c4 + + z11 = tmp7 + z3; // phase 5 + z13 = tmp7 - z3; + + *d5p = z13 + z2; // phase 6 + *d3p = z13 - z2; + *d1p = z11 + z4; + *d7p = z11 - z4; + + *d0p = d0; *d2p = d2; *d4p = d4; *d6p = d6; +} + +static void stbiw__jpg_calcBits(int val, unsigned short bits[2]) { + int tmp1 = val < 0 ? -val : val; + val = val < 0 ? val-1 : val; + bits[1] = 1; + while(tmp1 >>= 1) { + ++bits[1]; + } + bits[0] = val & ((1<0)&&(DU[end0pos]==0); --end0pos) { + } + // end0pos = first element in reverse order !=0 + if(end0pos == 0) { + stbiw__jpg_writeBits(s, bitBuf, bitCnt, EOB); + return DU[0]; + } + for(i = 1; i <= end0pos; ++i) { + int startpos = i; + int nrzeroes; + unsigned short bits[2]; + for (; DU[i]==0 && i<=end0pos; ++i) { + } + nrzeroes = i-startpos; + if ( nrzeroes >= 16 ) { + int lng = nrzeroes>>4; + int nrmarker; + for (nrmarker=1; nrmarker <= lng; ++nrmarker) +stbiw__jpg_writeBits(s, bitBuf, bitCnt, M16zeroes); + nrzeroes &= 15; + } + stbiw__jpg_calcBits(DU[i], bits); + stbiw__jpg_writeBits(s, bitBuf, bitCnt, HTAC[(nrzeroes<<4)+bits[1]]); + stbiw__jpg_writeBits(s, bitBuf, bitCnt, bits); + } + if(end0pos != 63) { + stbiw__jpg_writeBits(s, bitBuf, bitCnt, EOB); + } + return DU[0]; +} + +static int stbi_write_jpg_core(stbi__write_context *s, int width, int height, int comp, const void* data, int quality) { + // Constants that don't pollute global namespace + static const unsigned char std_dc_luminance_nrcodes[] = {0,0,1,5,1,1,1,1,1,1,0,0,0,0,0,0,0}; + static const unsigned char std_dc_luminance_values[] = {0,1,2,3,4,5,6,7,8,9,10,11}; + static const unsigned char std_ac_luminance_nrcodes[] = {0,0,2,1,3,3,2,4,3,5,5,4,4,0,0,1,0x7d}; + static const unsigned char std_ac_luminance_values[] = { + 0x01,0x02,0x03,0x00,0x04,0x11,0x05,0x12,0x21,0x31,0x41,0x06,0x13,0x51,0x61,0x07,0x22,0x71,0x14,0x32,0x81,0x91,0xa1,0x08, + 0x23,0x42,0xb1,0xc1,0x15,0x52,0xd1,0xf0,0x24,0x33,0x62,0x72,0x82,0x09,0x0a,0x16,0x17,0x18,0x19,0x1a,0x25,0x26,0x27,0x28, + 0x29,0x2a,0x34,0x35,0x36,0x37,0x38,0x39,0x3a,0x43,0x44,0x45,0x46,0x47,0x48,0x49,0x4a,0x53,0x54,0x55,0x56,0x57,0x58,0x59, + 0x5a,0x63,0x64,0x65,0x66,0x67,0x68,0x69,0x6a,0x73,0x74,0x75,0x76,0x77,0x78,0x79,0x7a,0x83,0x84,0x85,0x86,0x87,0x88,0x89, + 0x8a,0x92,0x93,0x94,0x95,0x96,0x97,0x98,0x99,0x9a,0xa2,0xa3,0xa4,0xa5,0xa6,0xa7,0xa8,0xa9,0xaa,0xb2,0xb3,0xb4,0xb5,0xb6, + 0xb7,0xb8,0xb9,0xba,0xc2,0xc3,0xc4,0xc5,0xc6,0xc7,0xc8,0xc9,0xca,0xd2,0xd3,0xd4,0xd5,0xd6,0xd7,0xd8,0xd9,0xda,0xe1,0xe2, + 0xe3,0xe4,0xe5,0xe6,0xe7,0xe8,0xe9,0xea,0xf1,0xf2,0xf3,0xf4,0xf5,0xf6,0xf7,0xf8,0xf9,0xfa + }; + static const unsigned char std_dc_chrominance_nrcodes[] = {0,0,3,1,1,1,1,1,1,1,1,1,0,0,0,0,0}; + static const unsigned char std_dc_chrominance_values[] = {0,1,2,3,4,5,6,7,8,9,10,11}; + static const unsigned char std_ac_chrominance_nrcodes[] = {0,0,2,1,2,4,4,3,4,7,5,4,4,0,1,2,0x77}; + static const unsigned char std_ac_chrominance_values[] = { + 0x00,0x01,0x02,0x03,0x11,0x04,0x05,0x21,0x31,0x06,0x12,0x41,0x51,0x07,0x61,0x71,0x13,0x22,0x32,0x81,0x08,0x14,0x42,0x91, + 0xa1,0xb1,0xc1,0x09,0x23,0x33,0x52,0xf0,0x15,0x62,0x72,0xd1,0x0a,0x16,0x24,0x34,0xe1,0x25,0xf1,0x17,0x18,0x19,0x1a,0x26, + 0x27,0x28,0x29,0x2a,0x35,0x36,0x37,0x38,0x39,0x3a,0x43,0x44,0x45,0x46,0x47,0x48,0x49,0x4a,0x53,0x54,0x55,0x56,0x57,0x58, + 0x59,0x5a,0x63,0x64,0x65,0x66,0x67,0x68,0x69,0x6a,0x73,0x74,0x75,0x76,0x77,0x78,0x79,0x7a,0x82,0x83,0x84,0x85,0x86,0x87, + 0x88,0x89,0x8a,0x92,0x93,0x94,0x95,0x96,0x97,0x98,0x99,0x9a,0xa2,0xa3,0xa4,0xa5,0xa6,0xa7,0xa8,0xa9,0xaa,0xb2,0xb3,0xb4, + 0xb5,0xb6,0xb7,0xb8,0xb9,0xba,0xc2,0xc3,0xc4,0xc5,0xc6,0xc7,0xc8,0xc9,0xca,0xd2,0xd3,0xd4,0xd5,0xd6,0xd7,0xd8,0xd9,0xda, + 0xe2,0xe3,0xe4,0xe5,0xe6,0xe7,0xe8,0xe9,0xea,0xf2,0xf3,0xf4,0xf5,0xf6,0xf7,0xf8,0xf9,0xfa + }; + // Huffman tables + static const unsigned short YDC_HT[256][2] = { {0,2},{2,3},{3,3},{4,3},{5,3},{6,3},{14,4},{30,5},{62,6},{126,7},{254,8},{510,9}}; + static const unsigned short UVDC_HT[256][2] = { {0,2},{1,2},{2,2},{6,3},{14,4},{30,5},{62,6},{126,7},{254,8},{510,9},{1022,10},{2046,11}}; + static const unsigned short YAC_HT[256][2] = { + {10,4},{0,2},{1,2},{4,3},{11,4},{26,5},{120,7},{248,8},{1014,10},{65410,16},{65411,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {12,4},{27,5},{121,7},{502,9},{2038,11},{65412,16},{65413,16},{65414,16},{65415,16},{65416,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {28,5},{249,8},{1015,10},{4084,12},{65417,16},{65418,16},{65419,16},{65420,16},{65421,16},{65422,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {58,6},{503,9},{4085,12},{65423,16},{65424,16},{65425,16},{65426,16},{65427,16},{65428,16},{65429,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {59,6},{1016,10},{65430,16},{65431,16},{65432,16},{65433,16},{65434,16},{65435,16},{65436,16},{65437,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {122,7},{2039,11},{65438,16},{65439,16},{65440,16},{65441,16},{65442,16},{65443,16},{65444,16},{65445,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {123,7},{4086,12},{65446,16},{65447,16},{65448,16},{65449,16},{65450,16},{65451,16},{65452,16},{65453,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {250,8},{4087,12},{65454,16},{65455,16},{65456,16},{65457,16},{65458,16},{65459,16},{65460,16},{65461,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {504,9},{32704,15},{65462,16},{65463,16},{65464,16},{65465,16},{65466,16},{65467,16},{65468,16},{65469,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {505,9},{65470,16},{65471,16},{65472,16},{65473,16},{65474,16},{65475,16},{65476,16},{65477,16},{65478,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {506,9},{65479,16},{65480,16},{65481,16},{65482,16},{65483,16},{65484,16},{65485,16},{65486,16},{65487,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {1017,10},{65488,16},{65489,16},{65490,16},{65491,16},{65492,16},{65493,16},{65494,16},{65495,16},{65496,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {1018,10},{65497,16},{65498,16},{65499,16},{65500,16},{65501,16},{65502,16},{65503,16},{65504,16},{65505,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {2040,11},{65506,16},{65507,16},{65508,16},{65509,16},{65510,16},{65511,16},{65512,16},{65513,16},{65514,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {65515,16},{65516,16},{65517,16},{65518,16},{65519,16},{65520,16},{65521,16},{65522,16},{65523,16},{65524,16},{0,0},{0,0},{0,0},{0,0},{0,0}, + {2041,11},{65525,16},{65526,16},{65527,16},{65528,16},{65529,16},{65530,16},{65531,16},{65532,16},{65533,16},{65534,16},{0,0},{0,0},{0,0},{0,0},{0,0} + }; + static const unsigned short UVAC_HT[256][2] = { + {0,2},{1,2},{4,3},{10,4},{24,5},{25,5},{56,6},{120,7},{500,9},{1014,10},{4084,12},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {11,4},{57,6},{246,8},{501,9},{2038,11},{4085,12},{65416,16},{65417,16},{65418,16},{65419,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {26,5},{247,8},{1015,10},{4086,12},{32706,15},{65420,16},{65421,16},{65422,16},{65423,16},{65424,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {27,5},{248,8},{1016,10},{4087,12},{65425,16},{65426,16},{65427,16},{65428,16},{65429,16},{65430,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {58,6},{502,9},{65431,16},{65432,16},{65433,16},{65434,16},{65435,16},{65436,16},{65437,16},{65438,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {59,6},{1017,10},{65439,16},{65440,16},{65441,16},{65442,16},{65443,16},{65444,16},{65445,16},{65446,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {121,7},{2039,11},{65447,16},{65448,16},{65449,16},{65450,16},{65451,16},{65452,16},{65453,16},{65454,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {122,7},{2040,11},{65455,16},{65456,16},{65457,16},{65458,16},{65459,16},{65460,16},{65461,16},{65462,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {249,8},{65463,16},{65464,16},{65465,16},{65466,16},{65467,16},{65468,16},{65469,16},{65470,16},{65471,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {503,9},{65472,16},{65473,16},{65474,16},{65475,16},{65476,16},{65477,16},{65478,16},{65479,16},{65480,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {504,9},{65481,16},{65482,16},{65483,16},{65484,16},{65485,16},{65486,16},{65487,16},{65488,16},{65489,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {505,9},{65490,16},{65491,16},{65492,16},{65493,16},{65494,16},{65495,16},{65496,16},{65497,16},{65498,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {506,9},{65499,16},{65500,16},{65501,16},{65502,16},{65503,16},{65504,16},{65505,16},{65506,16},{65507,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {2041,11},{65508,16},{65509,16},{65510,16},{65511,16},{65512,16},{65513,16},{65514,16},{65515,16},{65516,16},{0,0},{0,0},{0,0},{0,0},{0,0},{0,0}, + {16352,14},{65517,16},{65518,16},{65519,16},{65520,16},{65521,16},{65522,16},{65523,16},{65524,16},{65525,16},{0,0},{0,0},{0,0},{0,0},{0,0}, + {1018,10},{32707,15},{65526,16},{65527,16},{65528,16},{65529,16},{65530,16},{65531,16},{65532,16},{65533,16},{65534,16},{0,0},{0,0},{0,0},{0,0},{0,0} + }; + static const int YQT[] = {16,11,10,16,24,40,51,61,12,12,14,19,26,58,60,55,14,13,16,24,40,57,69,56,14,17,22,29,51,87,80,62,18,22, + 37,56,68,109,103,77,24,35,55,64,81,104,113,92,49,64,78,87,103,121,120,101,72,92,95,98,112,100,103,99}; + static const int UVQT[] = {17,18,24,47,99,99,99,99,18,21,26,66,99,99,99,99,24,26,56,99,99,99,99,99,47,66,99,99,99,99,99,99, + 99,99,99,99,99,99,99,99,99,99,99,99,99,99,99,99,99,99,99,99,99,99,99,99,99,99,99,99,99,99,99,99}; + static const float aasf[] = { 1.0f * 2.828427125f, 1.387039845f * 2.828427125f, 1.306562965f * 2.828427125f, 1.175875602f * 2.828427125f, + 1.0f * 2.828427125f, 0.785694958f * 2.828427125f, 0.541196100f * 2.828427125f, 0.275899379f * 2.828427125f }; + + int row, col, i, k, subsample; + float fdtbl_Y[64], fdtbl_UV[64]; + unsigned char YTable[64], UVTable[64]; + + if(!data || !width || !height || comp > 4 || comp < 1) { + return 0; + } + + quality = quality ? quality : 90; + subsample = quality <= 90 ? 1 : 0; + quality = quality < 1 ? 1 : quality > 100 ? 100 : quality; + quality = quality < 50 ? 5000 / quality : 200 - quality * 2; + + for(i = 0; i < 64; ++i) { + int uvti, yti = (YQT[i]*quality+50)/100; + YTable[stbiw__jpg_ZigZag[i]] = (unsigned char) (yti < 1 ? 1 : yti > 255 ? 255 : yti); + uvti = (UVQT[i]*quality+50)/100; + UVTable[stbiw__jpg_ZigZag[i]] = (unsigned char) (uvti < 1 ? 1 : uvti > 255 ? 255 : uvti); + } + + for(row = 0, k = 0; row < 8; ++row) { + for(col = 0; col < 8; ++col, ++k) { + fdtbl_Y[k] = 1 / (YTable [stbiw__jpg_ZigZag[k]] * aasf[row] * aasf[col]); + fdtbl_UV[k] = 1 / (UVTable[stbiw__jpg_ZigZag[k]] * aasf[row] * aasf[col]); + } + } + + // Write Headers + { + static const unsigned char head0[] = { 0xFF,0xD8,0xFF,0xE0,0,0x10,'J','F','I','F',0,1,1,0,0,1,0,1,0,0,0xFF,0xDB,0,0x84,0 }; + static const unsigned char head2[] = { 0xFF,0xDA,0,0xC,3,1,0,2,0x11,3,0x11,0,0x3F,0 }; + const unsigned char head1[] = { 0xFF,0xC0,0,0x11,8,(unsigned char)(height>>8),STBIW_UCHAR(height),(unsigned char)(width>>8),STBIW_UCHAR(width), + 3,1,(unsigned char)(subsample?0x22:0x11),0,2,0x11,1,3,0x11,1,0xFF,0xC4,0x01,0xA2,0 }; + s->func(s->context, (void*)head0, sizeof(head0)); + s->func(s->context, (void*)YTable, sizeof(YTable)); + stbiw__putc(s, 1); + s->func(s->context, UVTable, sizeof(UVTable)); + s->func(s->context, (void*)head1, sizeof(head1)); + s->func(s->context, (void*)(std_dc_luminance_nrcodes+1), sizeof(std_dc_luminance_nrcodes)-1); + s->func(s->context, (void*)std_dc_luminance_values, sizeof(std_dc_luminance_values)); + stbiw__putc(s, 0x10); // HTYACinfo + s->func(s->context, (void*)(std_ac_luminance_nrcodes+1), sizeof(std_ac_luminance_nrcodes)-1); + s->func(s->context, (void*)std_ac_luminance_values, sizeof(std_ac_luminance_values)); + stbiw__putc(s, 1); // HTUDCinfo + s->func(s->context, (void*)(std_dc_chrominance_nrcodes+1), sizeof(std_dc_chrominance_nrcodes)-1); + s->func(s->context, (void*)std_dc_chrominance_values, sizeof(std_dc_chrominance_values)); + stbiw__putc(s, 0x11); // HTUACinfo + s->func(s->context, (void*)(std_ac_chrominance_nrcodes+1), sizeof(std_ac_chrominance_nrcodes)-1); + s->func(s->context, (void*)std_ac_chrominance_values, sizeof(std_ac_chrominance_values)); + s->func(s->context, (void*)head2, sizeof(head2)); + } + + // Encode 8x8 macroblocks + { + static const unsigned short fillBits[] = {0x7F, 7}; + int DCY=0, DCU=0, DCV=0; + int bitBuf=0, bitCnt=0; + // comp == 2 is grey+alpha (alpha is ignored) + int ofsG = comp > 2 ? 1 : 0, ofsB = comp > 2 ? 2 : 0; + const unsigned char *dataR = (const unsigned char *)data; + const unsigned char *dataG = dataR + ofsG; + const unsigned char *dataB = dataR + ofsB; + int x, y, pos; + if(subsample) { + for(y = 0; y < height; y += 16) { +for(x = 0; x < width; x += 16) { + float Y[256], U[256], V[256]; + for(row = y, pos = 0; row < y+16; ++row) { + // row >= height => use last input row + int clamped_row = (row < height) ? row : height - 1; + int base_p = (stbi__flip_vertically_on_write ? (height-1-clamped_row) : clamped_row)*width*comp; + for(col = x; col < x+16; ++col, ++pos) { + // if col >= width => use pixel from last input column + int p = base_p + ((col < width) ? col : (width-1))*comp; + float r = dataR[p], g = dataG[p], b = dataB[p]; + Y[pos]= +0.29900f*r + 0.58700f*g + 0.11400f*b - 128; + U[pos]= -0.16874f*r - 0.33126f*g + 0.50000f*b; + V[pos]= +0.50000f*r - 0.41869f*g - 0.08131f*b; + } + } + DCY = stbiw__jpg_processDU(s, &bitBuf, &bitCnt, Y+0, 16, fdtbl_Y, DCY, YDC_HT, YAC_HT); + DCY = stbiw__jpg_processDU(s, &bitBuf, &bitCnt, Y+8, 16, fdtbl_Y, DCY, YDC_HT, YAC_HT); + DCY = stbiw__jpg_processDU(s, &bitBuf, &bitCnt, Y+128, 16, fdtbl_Y, DCY, YDC_HT, YAC_HT); + DCY = stbiw__jpg_processDU(s, &bitBuf, &bitCnt, Y+136, 16, fdtbl_Y, DCY, YDC_HT, YAC_HT); + + // subsample U,V + { + float subU[64], subV[64]; + int yy, xx; + for(yy = 0, pos = 0; yy < 8; ++yy) { + for(xx = 0; xx < 8; ++xx, ++pos) { +int j = yy*32+xx*2; +subU[pos] = (U[j+0] + U[j+1] + U[j+16] + U[j+17]) * 0.25f; +subV[pos] = (V[j+0] + V[j+1] + V[j+16] + V[j+17]) * 0.25f; + } + } + DCU = stbiw__jpg_processDU(s, &bitBuf, &bitCnt, subU, 8, fdtbl_UV, DCU, UVDC_HT, UVAC_HT); + DCV = stbiw__jpg_processDU(s, &bitBuf, &bitCnt, subV, 8, fdtbl_UV, DCV, UVDC_HT, UVAC_HT); + } +} + } + } else { + for(y = 0; y < height; y += 8) { +for(x = 0; x < width; x += 8) { + float Y[64], U[64], V[64]; + for(row = y, pos = 0; row < y+8; ++row) { + // row >= height => use last input row + int clamped_row = (row < height) ? row : height - 1; + int base_p = (stbi__flip_vertically_on_write ? (height-1-clamped_row) : clamped_row)*width*comp; + for(col = x; col < x+8; ++col, ++pos) { + // if col >= width => use pixel from last input column + int p = base_p + ((col < width) ? col : (width-1))*comp; + float r = dataR[p], g = dataG[p], b = dataB[p]; + Y[pos]= +0.29900f*r + 0.58700f*g + 0.11400f*b - 128; + U[pos]= -0.16874f*r - 0.33126f*g + 0.50000f*b; + V[pos]= +0.50000f*r - 0.41869f*g - 0.08131f*b; + } + } + + DCY = stbiw__jpg_processDU(s, &bitBuf, &bitCnt, Y, 8, fdtbl_Y, DCY, YDC_HT, YAC_HT); + DCU = stbiw__jpg_processDU(s, &bitBuf, &bitCnt, U, 8, fdtbl_UV, DCU, UVDC_HT, UVAC_HT); + DCV = stbiw__jpg_processDU(s, &bitBuf, &bitCnt, V, 8, fdtbl_UV, DCV, UVDC_HT, UVAC_HT); +} + } + } + + // Do the bit alignment of the EOI marker + stbiw__jpg_writeBits(s, &bitBuf, &bitCnt, fillBits); + } + + // EOI + stbiw__putc(s, 0xFF); + stbiw__putc(s, 0xD9); + + return 1; +} + +STBIWDEF int stbi_write_jpg_to_func(stbi_write_func *func, void *context, int x, int y, int comp, const void *data, int quality) +{ + stbi__write_context s; + stbi__start_write_callbacks(&s, func, context); + return stbi_write_jpg_core(&s, x, y, comp, (void *) data, quality); +} + + +#ifndef STBI_WRITE_NO_STDIO +STBIWDEF int stbi_write_jpg(char const *filename, int x, int y, int comp, const void *data, int quality) +{ + stbi__write_context s; + if (stbi__start_write_file(&s,filename)) { + int r = stbi_write_jpg_core(&s, x, y, comp, data, quality); + stbi__end_write_file(&s); + return r; + } else + return 0; +} +#endif + +#endif // STB_IMAGE_WRITE_IMPLEMENTATION + +/* Revision history + 1.14 (2020-02-02) updated JPEG writer to downsample chroma channels + 1.13 + 1.12 + 1.11 (2019-08-11) + + 1.10 (2019-02-07) + support utf8 filenames in Windows; fix warnings and platform ifdefs + 1.09 (2018-02-11) + fix typo in zlib quality API, improve STB_I_W_STATIC in C++ + 1.08 (2018-01-29) + add stbi__flip_vertically_on_write, external zlib, zlib quality, choose PNG filter + 1.07 (2017-07-24) + doc fix + 1.06 (2017-07-23) + writing JPEG (using Jon Olick's code) + 1.05 ??? + 1.04 (2017-03-03) + monochrome BMP expansion + 1.03 ??? + 1.02 (2016-04-02) + avoid allocating large structures on the stack + 1.01 (2016-01-16) + STBIW_REALLOC_SIZED: support allocators with no realloc support + avoid race-condition in crc initialization + minor compile issues + 1.00 (2015-09-14) + installable file IO function + 0.99 (2015-09-13) + warning fixes; TGA rle support + 0.98 (2015-04-08) + added STBIW_MALLOC, STBIW_ASSERT etc + 0.97 (2015-01-18) + fixed HDR asserts, rewrote HDR rle logic + 0.96 (2015-01-17) + add HDR output + fix monochrome BMP + 0.95 (2014-08-17) + add monochrome TGA output + 0.94 (2014-05-31) + rename private functions to avoid conflicts with stb_image.h + 0.93 (2014-05-27) + warning fixes + 0.92 (2010-08-01) + casts to unsigned char to fix warnings + 0.91 (2010-07-17) + first public release + 0.90 first internal release +*/ + +/* +------------------------------------------------------------------------------ +This software is available under 2 licenses -- choose whichever you prefer. +------------------------------------------------------------------------------ +ALTERNATIVE A - MIT License +Copyright (c) 2017 Sean Barrett +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. +------------------------------------------------------------------------------ +ALTERNATIVE B - Public Domain (www.unlicense.org) +This is free and unencumbered software released into the public domain. +Anyone is free to copy, modify, publish, use, compile, sell, or distribute this +software, either in source code form or as a compiled binary, for any purpose, +commercial or non-commercial, and by any means. +In jurisdictions that recognize copyright laws, the author or authors of this +software dedicate any and all copyright interest in the software to the public +domain. We make this dedication for the benefit of the public at large and to +the detriment of our heirs and successors. We intend this dedication to be an +overt act of relinquishment in perpetuity of all present and future rights to +this software under copyright law. +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 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. +------------------------------------------------------------------------------ +*/ diff --git a/private/blah/third_party/stb_truetype.h b/private/blah/third_party/stb_truetype.h new file mode 100644 index 0000000..e52dbee --- /dev/null +++ b/private/blah/third_party/stb_truetype.h @@ -0,0 +1,5011 @@ +// stb_truetype.h - v1.24 - public domain +// authored from 2009-2020 by Sean Barrett / RAD Game Tools +// +// ======================================================================= +// +// NO SECURITY GUARANTEE -- DO NOT USE THIS ON UNTRUSTED FONT FILES +// +// This library does no range checking of the offsets found in the file, +// meaning an attacker can use it to read arbitrary memory. +// +// ======================================================================= +// +// This library processes TrueType files: +// parse files +// extract glyph metrics +// extract glyph shapes +// render glyphs to one-channel bitmaps with antialiasing (box filter) +// render glyphs to one-channel SDF bitmaps (signed-distance field/function) +// +// Todo: +// non-MS cmaps +// crashproof on bad data +// hinting? (no longer patented) +// cleartype-style AA? +// optimize: use simple memory allocator for intermediates +// optimize: build edge-list directly from curves +// optimize: rasterize directly from curves? +// +// ADDITIONAL CONTRIBUTORS +// +// Mikko Mononen: compound shape support, more cmap formats +// Tor Andersson: kerning, subpixel rendering +// Dougall Johnson: OpenType / Type 2 font handling +// Daniel Ribeiro Maciel: basic GPOS-based kerning +// +// Misc other: +// Ryan Gordon +// Simon Glass +// github:IntellectualKitty +// Imanol Celaya +// Daniel Ribeiro Maciel +// +// Bug/warning reports/fixes: +// "Zer" on mollyrocket Fabian "ryg" Giesen github:NiLuJe +// Cass Everitt Martins Mozeiko github:aloucks +// stoiko (Haemimont Games) Cap Petschulat github:oyvindjam +// Brian Hook Omar Cornut github:vassvik +// Walter van Niftrik Ryan Griege +// David Gow Peter LaValle +// David Given Sergey Popov +// Ivan-Assen Ivanov Giumo X. Clanjor +// Anthony Pesch Higor Euripedes +// Johan Duparc Thomas Fields +// Hou Qiming Derek Vinyard +// Rob Loach Cort Stratton +// Kenney Phillis Jr. Brian Costabile +// Ken Voskuil (kaesve) +// +// VERSION HISTORY +// +// 1.24 (2020-02-05) fix warning +// 1.23 (2020-02-02) query SVG data for glyphs; query whole kerning table (but only kern not GPOS) +// 1.22 (2019-08-11) minimize missing-glyph duplication; fix kerning if both 'GPOS' and 'kern' are defined +// 1.21 (2019-02-25) fix warning +// 1.20 (2019-02-07) PackFontRange skips missing codepoints; GetScaleFontVMetrics() +// 1.19 (2018-02-11) GPOS kerning, STBTT_fmod +// 1.18 (2018-01-29) add missing function +// 1.17 (2017-07-23) make more arguments const; doc fix +// 1.16 (2017-07-12) SDF support +// 1.15 (2017-03-03) make more arguments const +// 1.14 (2017-01-16) num-fonts-in-TTC function +// 1.13 (2017-01-02) support OpenType fonts, certain Apple fonts +// 1.12 (2016-10-25) suppress warnings about casting away const with -Wcast-qual +// 1.11 (2016-04-02) fix unused-variable warning +// 1.10 (2016-04-02) user-defined fabs(); rare memory leak; remove duplicate typedef +// 1.09 (2016-01-16) warning fix; avoid crash on outofmem; use allocation userdata properly +// 1.08 (2015-09-13) document stbtt_Rasterize(); fixes for vertical & horizontal edges +// 1.07 (2015-08-01) allow PackFontRanges to accept arrays of sparse codepoints; +// variant PackFontRanges to pack and render in separate phases; +// fix stbtt_GetFontOFfsetForIndex (never worked for non-0 input?); +// fixed an assert() bug in the new rasterizer +// replace assert() with STBTT_assert() in new rasterizer +// +// Full history can be found at the end of this file. +// +// LICENSE +// +// See end of file for license information. +// +// USAGE +// +// Include this file in whatever places need to refer to it. In ONE C/C++ +// file, write: +// #define STB_TRUETYPE_IMPLEMENTATION +// before the #include of this file. This expands out the actual +// implementation into that C/C++ file. +// +// To make the implementation private to the file that generates the implementation, +// #define STBTT_STATIC +// +// Simple 3D API (don't ship this, but it's fine for tools and quick start) +// stbtt_BakeFontBitmap() -- bake a font to a bitmap for use as texture +// stbtt_GetBakedQuad() -- compute quad to draw for a given char +// +// Improved 3D API (more shippable): +// #include "stb_rect_pack.h" -- optional, but you really want it +// stbtt_PackBegin() +// stbtt_PackSetOversampling() -- for improved quality on small fonts +// stbtt_PackFontRanges() -- pack and renders +// stbtt_PackEnd() +// stbtt_GetPackedQuad() +// +// "Load" a font file from a memory buffer (you have to keep the buffer loaded) +// stbtt_InitFont() +// stbtt_GetFontOffsetForIndex() -- indexing for TTC font collections +// stbtt_GetNumberOfFonts() -- number of fonts for TTC font collections +// +// Render a unicode codepoint to a bitmap +// stbtt_GetCodepointBitmap() -- allocates and returns a bitmap +// stbtt_MakeCodepointBitmap() -- renders into bitmap you provide +// stbtt_GetCodepointBitmapBox() -- how big the bitmap must be +// +// Character advance/positioning +// stbtt_GetCodepointHMetrics() +// stbtt_GetFontVMetrics() +// stbtt_GetFontVMetricsOS2() +// stbtt_GetCodepointKernAdvance() +// +// Starting with version 1.06, the rasterizer was replaced with a new, +// faster and generally-more-precise rasterizer. The new rasterizer more +// accurately measures pixel coverage for anti-aliasing, except in the case +// where multiple shapes overlap, in which case it overestimates the AA pixel +// coverage. Thus, anti-aliasing of intersecting shapes may look wrong. If +// this turns out to be a problem, you can re-enable the old rasterizer with +// #define STBTT_RASTERIZER_VERSION 1 +// which will incur about a 15% speed hit. +// +// ADDITIONAL DOCUMENTATION +// +// Immediately after this block comment are a series of sample programs. +// +// After the sample programs is the "header file" section. This section +// includes documentation for each API function. +// +// Some important concepts to understand to use this library: +// +// Codepoint +// Characters are defined by unicode codepoints, e.g. 65 is +// uppercase A, 231 is lowercase c with a cedilla, 0x7e30 is +// the hiragana for "ma". +// +// Glyph +// A visual character shape (every codepoint is rendered as +// some glyph) +// +// Glyph index +// A font-specific integer ID representing a glyph +// +// Baseline +// Glyph shapes are defined relative to a baseline, which is the +// bottom of uppercase characters. Characters extend both above +// and below the baseline. +// +// Current Point +// As you draw text to the screen, you keep track of a "current point" +// which is the origin of each character. The current point's vertical +// position is the baseline. Even "baked fonts" use this model. +// +// Vertical Font Metrics +// The vertical qualities of the font, used to vertically position +// and space the characters. See docs for stbtt_GetFontVMetrics. +// +// Font Size in Pixels or Points +// The preferred interface for specifying font sizes in stb_truetype +// is to specify how tall the font's vertical extent should be in pixels. +// If that sounds good enough, skip the next paragraph. +// +// Most font APIs instead use "points", which are a common typographic +// measurement for describing font size, defined as 72 points per inch. +// stb_truetype provides a point API for compatibility. However, true +// "per inch" conventions don't make much sense on computer displays +// since different monitors have different number of pixels per +// inch. For example, Windows traditionally uses a convention that +// there are 96 pixels per inch, thus making 'inch' measurements have +// nothing to do with inches, and thus effectively defining a point to +// be 1.333 pixels. Additionally, the TrueType font data provides +// an explicit scale factor to scale a given font's glyphs to points, +// but the author has observed that this scale factor is often wrong +// for non-commercial fonts, thus making fonts scaled in points +// according to the TrueType spec incoherently sized in practice. +// +// DETAILED USAGE: +// +// Scale: +// Select how high you want the font to be, in points or pixels. +// Call ScaleForPixelHeight or ScaleForMappingEmToPixels to compute +// a scale factor SF that will be used by all other functions. +// +// Baseline: +// You need to select a y-coordinate that is the baseline of where +// your text will appear. Call GetFontBoundingBox to get the baseline-relative +// bounding box for all characters. SF*-y0 will be the distance in pixels +// that the worst-case character could extend above the baseline, so if +// you want the top edge of characters to appear at the top of the +// screen where y=0, then you would set the baseline to SF*-y0. +// +// Current point: +// Set the current point where the first character will appear. The +// first character could extend left of the current point; this is font +// dependent. You can either choose a current point that is the leftmost +// point and hope, or add some padding, or check the bounding box or +// left-side-bearing of the first character to be displayed and set +// the current point based on that. +// +// Displaying a character: +// Compute the bounding box of the character. It will contain signed values +// relative to . I.e. if it returns x0,y0,x1,y1, +// then the character should be displayed in the rectangle from +// to = 32 && *text < 128) { + stbtt_aligned_quad q; + stbtt_GetBakedQuad(cdata, 512,512, *text-32, &x,&y,&q,1);//1=opengl & d3d10+,0=d3d9 + glTexCoord2f(q.s0,q.t1); glVertex2f(q.x0,q.y0); + glTexCoord2f(q.s1,q.t1); glVertex2f(q.x1,q.y0); + glTexCoord2f(q.s1,q.t0); glVertex2f(q.x1,q.y1); + glTexCoord2f(q.s0,q.t0); glVertex2f(q.x0,q.y1); + } + ++text; + } + glEnd(); +} +#endif +// +// +////////////////////////////////////////////////////////////////////////////// +// +// Complete program (this compiles): get a single bitmap, print as ASCII art +// +#if 0 +#include +#define STB_TRUETYPE_IMPLEMENTATION // force following include to generate implementation +#include "stb_truetype.h" + +char ttf_buffer[1<<25]; + +int main(int argc, char **argv) +{ + stbtt_fontinfo font; + unsigned char *bitmap; + int w,h,i,j,c = (argc > 1 ? atoi(argv[1]) : 'a'), s = (argc > 2 ? atoi(argv[2]) : 20); + + fread(ttf_buffer, 1, 1<<25, fopen(argc > 3 ? argv[3] : "c:/windows/fonts/arialbd.ttf", "rb")); + + stbtt_InitFont(&font, ttf_buffer, stbtt_GetFontOffsetForIndex(ttf_buffer,0)); + bitmap = stbtt_GetCodepointBitmap(&font, 0,stbtt_ScaleForPixelHeight(&font, s), c, &w, &h, 0,0); + + for (j=0; j < h; ++j) { + for (i=0; i < w; ++i) + putchar(" .:ioVM@"[bitmap[j*w+i]>>5]); + putchar('\n'); + } + return 0; +} +#endif +// +// Output: +// +// .ii. +// @@@@@@. +// V@Mio@@o +// :i. V@V +// :oM@@M +// :@@@MM@M +// @@o o@M +// :@@. M@M +// @@@o@@@@ +// :M@@V:@@. +// +////////////////////////////////////////////////////////////////////////////// +// +// Complete program: print "Hello World!" banner, with bugs +// +#if 0 +char buffer[24<<20]; +unsigned char screen[20][79]; + +int main(int arg, char **argv) +{ + stbtt_fontinfo font; + int i,j,ascent,baseline,ch=0; + float scale, xpos=2; // leave a little padding in case the character extends left + char *text = "Heljo World!"; // intentionally misspelled to show 'lj' brokenness + + fread(buffer, 1, 1000000, fopen("c:/windows/fonts/arialbd.ttf", "rb")); + stbtt_InitFont(&font, buffer, 0); + + scale = stbtt_ScaleForPixelHeight(&font, 15); + stbtt_GetFontVMetrics(&font, &ascent,0,0); + baseline = (int) (ascent*scale); + + while (text[ch]) { + int advance,lsb,x0,y0,x1,y1; + float x_shift = xpos - (float) floor(xpos); + stbtt_GetCodepointHMetrics(&font, text[ch], &advance, &lsb); + stbtt_GetCodepointBitmapBoxSubpixel(&font, text[ch], scale,scale,x_shift,0, &x0,&y0,&x1,&y1); + stbtt_MakeCodepointBitmapSubpixel(&font, &screen[baseline + y0][(int) xpos + x0], x1-x0,y1-y0, 79, scale,scale,x_shift,0, text[ch]); + // note that this stomps the old data, so where character boxes overlap (e.g. 'lj') it's wrong + // because this API is really for baking character bitmaps into textures. if you want to render + // a sequence of characters, you really need to render each bitmap to a temp buffer, then + // "alpha blend" that into the working buffer + xpos += (advance * scale); + if (text[ch+1]) + xpos += scale*stbtt_GetCodepointKernAdvance(&font, text[ch],text[ch+1]); + ++ch; + } + + for (j=0; j < 20; ++j) { + for (i=0; i < 78; ++i) + putchar(" .:ioVM@"[screen[j][i]>>5]); + putchar('\n'); + } + + return 0; +} +#endif + + +////////////////////////////////////////////////////////////////////////////// +////////////////////////////////////////////////////////////////////////////// +//// +//// INTEGRATION WITH YOUR CODEBASE +//// +//// The following sections allow you to supply alternate definitions +//// of C library functions used by stb_truetype, e.g. if you don't +//// link with the C runtime library. + +#ifdef STB_TRUETYPE_IMPLEMENTATION + // #define your own (u)stbtt_int8/16/32 before including to override this + #ifndef stbtt_uint8 + typedef unsigned char stbtt_uint8; + typedef signed char stbtt_int8; + typedef unsigned short stbtt_uint16; + typedef signed short stbtt_int16; + typedef unsigned int stbtt_uint32; + typedef signed int stbtt_int32; + #endif + + typedef char stbtt__check_size32[sizeof(stbtt_int32)==4 ? 1 : -1]; + typedef char stbtt__check_size16[sizeof(stbtt_int16)==2 ? 1 : -1]; + + // e.g. #define your own STBTT_ifloor/STBTT_iceil() to avoid math.h + #ifndef STBTT_ifloor + #include + #define STBTT_ifloor(x) ((int) floor(x)) + #define STBTT_iceil(x) ((int) ceil(x)) + #endif + + #ifndef STBTT_sqrt + #include + #define STBTT_sqrt(x) sqrt(x) + #define STBTT_pow(x,y) pow(x,y) + #endif + + #ifndef STBTT_fmod + #include + #define STBTT_fmod(x,y) fmod(x,y) + #endif + + #ifndef STBTT_cos + #include + #define STBTT_cos(x) cos(x) + #define STBTT_acos(x) acos(x) + #endif + + #ifndef STBTT_fabs + #include + #define STBTT_fabs(x) fabs(x) + #endif + + // #define your own functions "STBTT_malloc" / "STBTT_free" to avoid malloc.h + #ifndef STBTT_malloc + #include + #define STBTT_malloc(x,u) ((void)(u),malloc(x)) + #define STBTT_free(x,u) ((void)(u),free(x)) + #endif + + #ifndef STBTT_assert + #include + #define STBTT_assert(x) assert(x) + #endif + + #ifndef STBTT_strlen + #include + #define STBTT_strlen(x) strlen(x) + #endif + + #ifndef STBTT_memcpy + #include + #define STBTT_memcpy memcpy + #define STBTT_memset memset + #endif +#endif + +/////////////////////////////////////////////////////////////////////////////// +/////////////////////////////////////////////////////////////////////////////// +//// +//// INTERFACE +//// +//// + +#ifndef __STB_INCLUDE_STB_TRUETYPE_H__ +#define __STB_INCLUDE_STB_TRUETYPE_H__ + +#ifdef STBTT_STATIC +#define STBTT_DEF static +#else +#define STBTT_DEF extern +#endif + +#ifdef __cplusplus +extern "C" { +#endif + +// private structure +typedef struct +{ + unsigned char *data; + int cursor; + int size; +} stbtt__buf; + +////////////////////////////////////////////////////////////////////////////// +// +// TEXTURE BAKING API +// +// If you use this API, you only have to call two functions ever. +// + +typedef struct +{ + unsigned short x0,y0,x1,y1; // coordinates of bbox in bitmap + float xoff,yoff,xadvance; +} stbtt_bakedchar; + +STBTT_DEF int stbtt_BakeFontBitmap(const unsigned char *data, int offset, // font location (use offset=0 for plain .ttf) + float pixel_height, // height of font in pixels + unsigned char *pixels, int pw, int ph, // bitmap to be filled in + int first_char, int num_chars, // characters to bake + stbtt_bakedchar *chardata); // you allocate this, it's num_chars long +// if return is positive, the first unused row of the bitmap +// if return is negative, returns the negative of the number of characters that fit +// if return is 0, no characters fit and no rows were used +// This uses a very crappy packing. + +typedef struct +{ + float x0,y0,s0,t0; // top-left + float x1,y1,s1,t1; // bottom-right +} stbtt_aligned_quad; + +STBTT_DEF void stbtt_GetBakedQuad(const stbtt_bakedchar *chardata, int pw, int ph, // same data as above + int char_index, // character to display + float *xpos, float *ypos, // pointers to current position in screen pixel space + stbtt_aligned_quad *q, // output: quad to draw + int opengl_fillrule); // true if opengl fill rule; false if DX9 or earlier +// Call GetBakedQuad with char_index = 'character - first_char', and it +// creates the quad you need to draw and advances the current position. +// +// The coordinate system used assumes y increases downwards. +// +// Characters will extend both above and below the current position; +// see discussion of "BASELINE" above. +// +// It's inefficient; you might want to c&p it and optimize it. + +STBTT_DEF void stbtt_GetScaledFontVMetrics(const unsigned char *fontdata, int index, float size, float *ascent, float *descent, float *lineGap); +// Query the font vertical metrics without having to create a font first. + + +////////////////////////////////////////////////////////////////////////////// +// +// NEW TEXTURE BAKING API +// +// This provides options for packing multiple fonts into one atlas, not +// perfectly but better than nothing. + +typedef struct +{ + unsigned short x0,y0,x1,y1; // coordinates of bbox in bitmap + float xoff,yoff,xadvance; + float xoff2,yoff2; +} stbtt_packedchar; + +typedef struct stbtt_pack_context stbtt_pack_context; +typedef struct stbtt_fontinfo stbtt_fontinfo; +#ifndef STB_RECT_PACK_VERSION +typedef struct stbrp_rect stbrp_rect; +#endif + +STBTT_DEF int stbtt_PackBegin(stbtt_pack_context *spc, unsigned char *pixels, int width, int height, int stride_in_bytes, int padding, void *alloc_context); +// Initializes a packing context stored in the passed-in stbtt_pack_context. +// Future calls using this context will pack characters into the bitmap passed +// in here: a 1-channel bitmap that is width * height. stride_in_bytes is +// the distance from one row to the next (or 0 to mean they are packed tightly +// together). "padding" is the amount of padding to leave between each +// character (normally you want '1' for bitmaps you'll use as textures with +// bilinear filtering). +// +// Returns 0 on failure, 1 on success. + +STBTT_DEF void stbtt_PackEnd (stbtt_pack_context *spc); +// Cleans up the packing context and frees all memory. + +#define STBTT_POINT_SIZE(x) (-(x)) + +STBTT_DEF int stbtt_PackFontRange(stbtt_pack_context *spc, const unsigned char *fontdata, int font_index, float font_size, + int first_unicode_char_in_range, int num_chars_in_range, stbtt_packedchar *chardata_for_range); +// Creates character bitmaps from the font_index'th font found in fontdata (use +// font_index=0 if you don't know what that is). It creates num_chars_in_range +// bitmaps for characters with unicode values starting at first_unicode_char_in_range +// and increasing. Data for how to render them is stored in chardata_for_range; +// pass these to stbtt_GetPackedQuad to get back renderable quads. +// +// font_size is the full height of the character from ascender to descender, +// as computed by stbtt_ScaleForPixelHeight. To use a point size as computed +// by stbtt_ScaleForMappingEmToPixels, wrap the point size in STBTT_POINT_SIZE() +// and pass that result as 'font_size': +// ..., 20 , ... // font max minus min y is 20 pixels tall +// ..., STBTT_POINT_SIZE(20), ... // 'M' is 20 pixels tall + +typedef struct +{ + float font_size; + int first_unicode_codepoint_in_range; // if non-zero, then the chars are continuous, and this is the first codepoint + int *array_of_unicode_codepoints; // if non-zero, then this is an array of unicode codepoints + int num_chars; + stbtt_packedchar *chardata_for_range; // output + unsigned char h_oversample, v_oversample; // don't set these, they're used internally +} stbtt_pack_range; + +STBTT_DEF int stbtt_PackFontRanges(stbtt_pack_context *spc, const unsigned char *fontdata, int font_index, stbtt_pack_range *ranges, int num_ranges); +// Creates character bitmaps from multiple ranges of characters stored in +// ranges. This will usually create a better-packed bitmap than multiple +// calls to stbtt_PackFontRange. Note that you can call this multiple +// times within a single PackBegin/PackEnd. + +STBTT_DEF void stbtt_PackSetOversampling(stbtt_pack_context *spc, unsigned int h_oversample, unsigned int v_oversample); +// Oversampling a font increases the quality by allowing higher-quality subpixel +// positioning, and is especially valuable at smaller text sizes. +// +// This function sets the amount of oversampling for all following calls to +// stbtt_PackFontRange(s) or stbtt_PackFontRangesGatherRects for a given +// pack context. The default (no oversampling) is achieved by h_oversample=1 +// and v_oversample=1. The total number of pixels required is +// h_oversample*v_oversample larger than the default; for example, 2x2 +// oversampling requires 4x the storage of 1x1. For best results, render +// oversampled textures with bilinear filtering. Look at the readme in +// stb/tests/oversample for information about oversampled fonts +// +// To use with PackFontRangesGather etc., you must set it before calls +// call to PackFontRangesGatherRects. + +STBTT_DEF void stbtt_PackSetSkipMissingCodepoints(stbtt_pack_context *spc, int skip); +// If skip != 0, this tells stb_truetype to skip any codepoints for which +// there is no corresponding glyph. If skip=0, which is the default, then +// codepoints without a glyph recived the font's "missing character" glyph, +// typically an empty box by convention. + +STBTT_DEF void stbtt_GetPackedQuad(const stbtt_packedchar *chardata, int pw, int ph, // same data as above + int char_index, // character to display + float *xpos, float *ypos, // pointers to current position in screen pixel space + stbtt_aligned_quad *q, // output: quad to draw + int align_to_integer); + +STBTT_DEF int stbtt_PackFontRangesGatherRects(stbtt_pack_context *spc, const stbtt_fontinfo *info, stbtt_pack_range *ranges, int num_ranges, stbrp_rect *rects); +STBTT_DEF void stbtt_PackFontRangesPackRects(stbtt_pack_context *spc, stbrp_rect *rects, int num_rects); +STBTT_DEF int stbtt_PackFontRangesRenderIntoRects(stbtt_pack_context *spc, const stbtt_fontinfo *info, stbtt_pack_range *ranges, int num_ranges, stbrp_rect *rects); +// Calling these functions in sequence is roughly equivalent to calling +// stbtt_PackFontRanges(). If you more control over the packing of multiple +// fonts, or if you want to pack custom data into a font texture, take a look +// at the source to of stbtt_PackFontRanges() and create a custom version +// using these functions, e.g. call GatherRects multiple times, +// building up a single array of rects, then call PackRects once, +// then call RenderIntoRects repeatedly. This may result in a +// better packing than calling PackFontRanges multiple times +// (or it may not). + +// this is an opaque structure that you shouldn't mess with which holds +// all the context needed from PackBegin to PackEnd. +struct stbtt_pack_context { + void *user_allocator_context; + void *pack_info; + int width; + int height; + int stride_in_bytes; + int padding; + int skip_missing; + unsigned int h_oversample, v_oversample; + unsigned char *pixels; + void *nodes; +}; + +////////////////////////////////////////////////////////////////////////////// +// +// FONT LOADING +// +// + +STBTT_DEF int stbtt_GetNumberOfFonts(const unsigned char *data); +// This function will determine the number of fonts in a font file. TrueType +// collection (.ttc) files may contain multiple fonts, while TrueType font +// (.ttf) files only contain one font. The number of fonts can be used for +// indexing with the previous function where the index is between zero and one +// less than the total fonts. If an error occurs, -1 is returned. + +STBTT_DEF int stbtt_GetFontOffsetForIndex(const unsigned char *data, int index); +// Each .ttf/.ttc file may have more than one font. Each font has a sequential +// index number starting from 0. Call this function to get the font offset for +// a given index; it returns -1 if the index is out of range. A regular .ttf +// file will only define one font and it always be at offset 0, so it will +// return '0' for index 0, and -1 for all other indices. + +// The following structure is defined publicly so you can declare one on +// the stack or as a global or etc, but you should treat it as opaque. +struct stbtt_fontinfo +{ + void * userdata; + unsigned char * data; // pointer to .ttf file + int fontstart; // offset of start of font + + int numGlyphs; // number of glyphs, needed for range checking + + int loca,head,glyf,hhea,hmtx,kern,gpos,svg; // table locations as offset from start of .ttf + int index_map; // a cmap mapping for our chosen character encoding + int indexToLocFormat; // format needed to map from glyph index to glyph + + stbtt__buf cff; // cff font data + stbtt__buf charstrings; // the charstring index + stbtt__buf gsubrs; // global charstring subroutines index + stbtt__buf subrs; // private charstring subroutines index + stbtt__buf fontdicts; // array of font dicts + stbtt__buf fdselect; // map from glyph to fontdict +}; + +STBTT_DEF int stbtt_InitFont(stbtt_fontinfo *info, const unsigned char *data, int offset); +// Given an offset into the file that defines a font, this function builds +// the necessary cached info for the rest of the system. You must allocate +// the stbtt_fontinfo yourself, and stbtt_InitFont will fill it out. You don't +// need to do anything special to free it, because the contents are pure +// value data with no additional data structures. Returns 0 on failure. + + +////////////////////////////////////////////////////////////////////////////// +// +// CHARACTER TO GLYPH-INDEX CONVERSIOn + +STBTT_DEF int stbtt_FindGlyphIndex(const stbtt_fontinfo *info, int unicode_codepoint); +// If you're going to perform multiple operations on the same character +// and you want a speed-up, call this function with the character you're +// going to process, then use glyph-based functions instead of the +// codepoint-based functions. +// Returns 0 if the character codepoint is not defined in the font. + + +////////////////////////////////////////////////////////////////////////////// +// +// CHARACTER PROPERTIES +// + +STBTT_DEF float stbtt_ScaleForPixelHeight(const stbtt_fontinfo *info, float pixels); +// computes a scale factor to produce a font whose "height" is 'pixels' tall. +// Height is measured as the distance from the highest ascender to the lowest +// descender; in other words, it's equivalent to calling stbtt_GetFontVMetrics +// and computing: +// scale = pixels / (ascent - descent) +// so if you prefer to measure height by the ascent only, use a similar calculation. + +STBTT_DEF float stbtt_ScaleForMappingEmToPixels(const stbtt_fontinfo *info, float pixels); +// computes a scale factor to produce a font whose EM size is mapped to +// 'pixels' tall. This is probably what traditional APIs compute, but +// I'm not positive. + +STBTT_DEF void stbtt_GetFontVMetrics(const stbtt_fontinfo *info, int *ascent, int *descent, int *lineGap); +// ascent is the coordinate above the baseline the font extends; descent +// is the coordinate below the baseline the font extends (i.e. it is typically negative) +// lineGap is the spacing between one row's descent and the next row's ascent... +// so you should advance the vertical position by "*ascent - *descent + *lineGap" +// these are expressed in unscaled coordinates, so you must multiply by +// the scale factor for a given size + +STBTT_DEF int stbtt_GetFontVMetricsOS2(const stbtt_fontinfo *info, int *typoAscent, int *typoDescent, int *typoLineGap); +// analogous to GetFontVMetrics, but returns the "typographic" values from the OS/2 +// table (specific to MS/Windows TTF files). +// +// Returns 1 on success (table present), 0 on failure. + +STBTT_DEF void stbtt_GetFontBoundingBox(const stbtt_fontinfo *info, int *x0, int *y0, int *x1, int *y1); +// the bounding box around all possible characters + +STBTT_DEF void stbtt_GetCodepointHMetrics(const stbtt_fontinfo *info, int codepoint, int *advanceWidth, int *leftSideBearing); +// leftSideBearing is the offset from the current horizontal position to the left edge of the character +// advanceWidth is the offset from the current horizontal position to the next horizontal position +// these are expressed in unscaled coordinates + +STBTT_DEF int stbtt_GetCodepointKernAdvance(const stbtt_fontinfo *info, int ch1, int ch2); +// an additional amount to add to the 'advance' value between ch1 and ch2 + +STBTT_DEF int stbtt_GetCodepointBox(const stbtt_fontinfo *info, int codepoint, int *x0, int *y0, int *x1, int *y1); +// Gets the bounding box of the visible part of the glyph, in unscaled coordinates + +STBTT_DEF void stbtt_GetGlyphHMetrics(const stbtt_fontinfo *info, int glyph_index, int *advanceWidth, int *leftSideBearing); +STBTT_DEF int stbtt_GetGlyphKernAdvance(const stbtt_fontinfo *info, int glyph1, int glyph2); +STBTT_DEF int stbtt_GetGlyphBox(const stbtt_fontinfo *info, int glyph_index, int *x0, int *y0, int *x1, int *y1); +// as above, but takes one or more glyph indices for greater efficiency + +typedef struct stbtt_kerningentry +{ + int glyph1; // use stbtt_FindGlyphIndex + int glyph2; + int advance; +} stbtt_kerningentry; + +STBTT_DEF int stbtt_GetKerningTableLength(const stbtt_fontinfo *info); +STBTT_DEF int stbtt_GetKerningTable(const stbtt_fontinfo *info, stbtt_kerningentry* table, int table_length); +// Retrieves a complete list of all of the kerning pairs provided by the font +// stbtt_GetKerningTable never writes more than table_length entries and returns how many entries it did write. +// The table will be sorted by (a.glyph1 == b.glyph1)?(a.glyph2 < b.glyph2):(a.glyph1 < b.glyph1) + +////////////////////////////////////////////////////////////////////////////// +// +// GLYPH SHAPES (you probably don't need these, but they have to go before +// the bitmaps for C declaration-order reasons) +// + +#ifndef STBTT_vmove // you can predefine these to use different values (but why?) + enum { + STBTT_vmove=1, + STBTT_vline, + STBTT_vcurve, + STBTT_vcubic + }; +#endif + +#ifndef stbtt_vertex // you can predefine this to use different values + // (we share this with other code at RAD) + #define stbtt_vertex_type short // can't use stbtt_int16 because that's not visible in the header file + typedef struct + { + stbtt_vertex_type x,y,cx,cy,cx1,cy1; + unsigned char type,padding; + } stbtt_vertex; +#endif + +STBTT_DEF int stbtt_IsGlyphEmpty(const stbtt_fontinfo *info, int glyph_index); +// returns non-zero if nothing is drawn for this glyph + +STBTT_DEF int stbtt_GetCodepointShape(const stbtt_fontinfo *info, int unicode_codepoint, stbtt_vertex **vertices); +STBTT_DEF int stbtt_GetGlyphShape(const stbtt_fontinfo *info, int glyph_index, stbtt_vertex **vertices); +// returns # of vertices and fills *vertices with the pointer to them +// these are expressed in "unscaled" coordinates +// +// The shape is a series of contours. Each one starts with +// a STBTT_moveto, then consists of a series of mixed +// STBTT_lineto and STBTT_curveto segments. A lineto +// draws a line from previous endpoint to its x,y; a curveto +// draws a quadratic bezier from previous endpoint to +// its x,y, using cx,cy as the bezier control point. + +STBTT_DEF void stbtt_FreeShape(const stbtt_fontinfo *info, stbtt_vertex *vertices); +// frees the data allocated above + +STBTT_DEF int stbtt_GetCodepointSVG(const stbtt_fontinfo *info, int unicode_codepoint, const char **svg); +STBTT_DEF int stbtt_GetGlyphSVG(const stbtt_fontinfo *info, int gl, const char **svg); +// fills svg with the character's SVG data. +// returns data size or 0 if SVG not found. + +////////////////////////////////////////////////////////////////////////////// +// +// BITMAP RENDERING +// + +STBTT_DEF void stbtt_FreeBitmap(unsigned char *bitmap, void *userdata); +// frees the bitmap allocated below + +STBTT_DEF unsigned char *stbtt_GetCodepointBitmap(const stbtt_fontinfo *info, float scale_x, float scale_y, int codepoint, int *width, int *height, int *xoff, int *yoff); +// allocates a large-enough single-channel 8bpp bitmap and renders the +// specified character/glyph at the specified scale into it, with +// antialiasing. 0 is no coverage (transparent), 255 is fully covered (opaque). +// *width & *height are filled out with the width & height of the bitmap, +// which is stored left-to-right, top-to-bottom. +// +// xoff/yoff are the offset it pixel space from the glyph origin to the top-left of the bitmap + +STBTT_DEF unsigned char *stbtt_GetCodepointBitmapSubpixel(const stbtt_fontinfo *info, float scale_x, float scale_y, float shift_x, float shift_y, int codepoint, int *width, int *height, int *xoff, int *yoff); +// the same as stbtt_GetCodepoitnBitmap, but you can specify a subpixel +// shift for the character + +STBTT_DEF void stbtt_MakeCodepointBitmap(const stbtt_fontinfo *info, unsigned char *output, int out_w, int out_h, int out_stride, float scale_x, float scale_y, int codepoint); +// the same as stbtt_GetCodepointBitmap, but you pass in storage for the bitmap +// in the form of 'output', with row spacing of 'out_stride' bytes. the bitmap +// is clipped to out_w/out_h bytes. Call stbtt_GetCodepointBitmapBox to get the +// width and height and positioning info for it first. + +STBTT_DEF void stbtt_MakeCodepointBitmapSubpixel(const stbtt_fontinfo *info, unsigned char *output, int out_w, int out_h, int out_stride, float scale_x, float scale_y, float shift_x, float shift_y, int codepoint); +// same as stbtt_MakeCodepointBitmap, but you can specify a subpixel +// shift for the character + +STBTT_DEF void stbtt_MakeCodepointBitmapSubpixelPrefilter(const stbtt_fontinfo *info, unsigned char *output, int out_w, int out_h, int out_stride, float scale_x, float scale_y, float shift_x, float shift_y, int oversample_x, int oversample_y, float *sub_x, float *sub_y, int codepoint); +// same as stbtt_MakeCodepointBitmapSubpixel, but prefiltering +// is performed (see stbtt_PackSetOversampling) + +STBTT_DEF void stbtt_GetCodepointBitmapBox(const stbtt_fontinfo *font, int codepoint, float scale_x, float scale_y, int *ix0, int *iy0, int *ix1, int *iy1); +// get the bbox of the bitmap centered around the glyph origin; so the +// bitmap width is ix1-ix0, height is iy1-iy0, and location to place +// the bitmap top left is (leftSideBearing*scale,iy0). +// (Note that the bitmap uses y-increases-down, but the shape uses +// y-increases-up, so CodepointBitmapBox and CodepointBox are inverted.) + +STBTT_DEF void stbtt_GetCodepointBitmapBoxSubpixel(const stbtt_fontinfo *font, int codepoint, float scale_x, float scale_y, float shift_x, float shift_y, int *ix0, int *iy0, int *ix1, int *iy1); +// same as stbtt_GetCodepointBitmapBox, but you can specify a subpixel +// shift for the character + +// the following functions are equivalent to the above functions, but operate +// on glyph indices instead of Unicode codepoints (for efficiency) +STBTT_DEF unsigned char *stbtt_GetGlyphBitmap(const stbtt_fontinfo *info, float scale_x, float scale_y, int glyph, int *width, int *height, int *xoff, int *yoff); +STBTT_DEF unsigned char *stbtt_GetGlyphBitmapSubpixel(const stbtt_fontinfo *info, float scale_x, float scale_y, float shift_x, float shift_y, int glyph, int *width, int *height, int *xoff, int *yoff); +STBTT_DEF void stbtt_MakeGlyphBitmap(const stbtt_fontinfo *info, unsigned char *output, int out_w, int out_h, int out_stride, float scale_x, float scale_y, int glyph); +STBTT_DEF void stbtt_MakeGlyphBitmapSubpixel(const stbtt_fontinfo *info, unsigned char *output, int out_w, int out_h, int out_stride, float scale_x, float scale_y, float shift_x, float shift_y, int glyph); +STBTT_DEF void stbtt_MakeGlyphBitmapSubpixelPrefilter(const stbtt_fontinfo *info, unsigned char *output, int out_w, int out_h, int out_stride, float scale_x, float scale_y, float shift_x, float shift_y, int oversample_x, int oversample_y, float *sub_x, float *sub_y, int glyph); +STBTT_DEF void stbtt_GetGlyphBitmapBox(const stbtt_fontinfo *font, int glyph, float scale_x, float scale_y, int *ix0, int *iy0, int *ix1, int *iy1); +STBTT_DEF void stbtt_GetGlyphBitmapBoxSubpixel(const stbtt_fontinfo *font, int glyph, float scale_x, float scale_y,float shift_x, float shift_y, int *ix0, int *iy0, int *ix1, int *iy1); + + +// @TODO: don't expose this structure +typedef struct +{ + int w,h,stride; + unsigned char *pixels; +} stbtt__bitmap; + +// rasterize a shape with quadratic beziers into a bitmap +STBTT_DEF void stbtt_Rasterize(stbtt__bitmap *result, // 1-channel bitmap to draw into + float flatness_in_pixels, // allowable error of curve in pixels + stbtt_vertex *vertices, // array of vertices defining shape + int num_verts, // number of vertices in above array + float scale_x, float scale_y, // scale applied to input vertices + float shift_x, float shift_y, // translation applied to input vertices + int x_off, int y_off, // another translation applied to input + int invert, // if non-zero, vertically flip shape + void *userdata); // context for to STBTT_MALLOC + +////////////////////////////////////////////////////////////////////////////// +// +// Signed Distance Function (or Field) rendering + +STBTT_DEF void stbtt_FreeSDF(unsigned char *bitmap, void *userdata); +// frees the SDF bitmap allocated below + +STBTT_DEF unsigned char * stbtt_GetGlyphSDF(const stbtt_fontinfo *info, float scale, int glyph, int padding, unsigned char onedge_value, float pixel_dist_scale, int *width, int *height, int *xoff, int *yoff); +STBTT_DEF unsigned char * stbtt_GetCodepointSDF(const stbtt_fontinfo *info, float scale, int codepoint, int padding, unsigned char onedge_value, float pixel_dist_scale, int *width, int *height, int *xoff, int *yoff); +// These functions compute a discretized SDF field for a single character, suitable for storing +// in a single-channel texture, sampling with bilinear filtering, and testing against +// larger than some threshold to produce scalable fonts. +// info -- the font +// scale -- controls the size of the resulting SDF bitmap, same as it would be creating a regular bitmap +// glyph/codepoint -- the character to generate the SDF for +// padding -- extra "pixels" around the character which are filled with the distance to the character (not 0), +// which allows effects like bit outlines +// onedge_value -- value 0-255 to test the SDF against to reconstruct the character (i.e. the isocontour of the character) +// pixel_dist_scale -- what value the SDF should increase by when moving one SDF "pixel" away from the edge (on the 0..255 scale) +// if positive, > onedge_value is inside; if negative, < onedge_value is inside +// width,height -- output height & width of the SDF bitmap (including padding) +// xoff,yoff -- output origin of the character +// return value -- a 2D array of bytes 0..255, width*height in size +// +// pixel_dist_scale & onedge_value are a scale & bias that allows you to make +// optimal use of the limited 0..255 for your application, trading off precision +// and special effects. SDF values outside the range 0..255 are clamped to 0..255. +// +// Example: +// scale = stbtt_ScaleForPixelHeight(22) +// padding = 5 +// onedge_value = 180 +// pixel_dist_scale = 180/5.0 = 36.0 +// +// This will create an SDF bitmap in which the character is about 22 pixels +// high but the whole bitmap is about 22+5+5=32 pixels high. To produce a filled +// shape, sample the SDF at each pixel and fill the pixel if the SDF value +// is greater than or equal to 180/255. (You'll actually want to antialias, +// which is beyond the scope of this example.) Additionally, you can compute +// offset outlines (e.g. to stroke the character border inside & outside, +// or only outside). For example, to fill outside the character up to 3 SDF +// pixels, you would compare against (180-36.0*3)/255 = 72/255. The above +// choice of variables maps a range from 5 pixels outside the shape to +// 2 pixels inside the shape to 0..255; this is intended primarily for apply +// outside effects only (the interior range is needed to allow proper +// antialiasing of the font at *smaller* sizes) +// +// The function computes the SDF analytically at each SDF pixel, not by e.g. +// building a higher-res bitmap and approximating it. In theory the quality +// should be as high as possible for an SDF of this size & representation, but +// unclear if this is true in practice (perhaps building a higher-res bitmap +// and computing from that can allow drop-out prevention). +// +// The algorithm has not been optimized at all, so expect it to be slow +// if computing lots of characters or very large sizes. + + + +////////////////////////////////////////////////////////////////////////////// +// +// Finding the right font... +// +// You should really just solve this offline, keep your own tables +// of what font is what, and don't try to get it out of the .ttf file. +// That's because getting it out of the .ttf file is really hard, because +// the names in the file can appear in many possible encodings, in many +// possible languages, and e.g. if you need a case-insensitive comparison, +// the details of that depend on the encoding & language in a complex way +// (actually underspecified in truetype, but also gigantic). +// +// But you can use the provided functions in two possible ways: +// stbtt_FindMatchingFont() will use *case-sensitive* comparisons on +// unicode-encoded names to try to find the font you want; +// you can run this before calling stbtt_InitFont() +// +// stbtt_GetFontNameStr() lets you get any of the various strings +// from the file yourself and do your own comparisons on them. +// You have to have called stbtt_InitFont() first. + + +STBTT_DEF int stbtt_FindMatchingFont(const unsigned char *fontdata, const char *name, int flags); +// returns the offset (not index) of the font that matches, or -1 if none +// if you use STBTT_MACSTYLE_DONTCARE, use a font name like "Arial Bold". +// if you use any other flag, use a font name like "Arial"; this checks +// the 'macStyle' header field; i don't know if fonts set this consistently +#define STBTT_MACSTYLE_DONTCARE 0 +#define STBTT_MACSTYLE_BOLD 1 +#define STBTT_MACSTYLE_ITALIC 2 +#define STBTT_MACSTYLE_UNDERSCORE 4 +#define STBTT_MACSTYLE_NONE 8 // <= not same as 0, this makes us check the bitfield is 0 + +STBTT_DEF int stbtt_CompareUTF8toUTF16_bigendian(const char *s1, int len1, const char *s2, int len2); +// returns 1/0 whether the first string interpreted as utf8 is identical to +// the second string interpreted as big-endian utf16... useful for strings from next func + +STBTT_DEF const char *stbtt_GetFontNameStr(const stbtt_fontinfo *font, int *length, int platformID, int encodingID, int languageID, int nameID); +// returns the string (which may be big-endian double byte, e.g. for unicode) +// and puts the length in bytes in *length. +// +// some of the values for the IDs are below; for more see the truetype spec: +// http://developer.apple.com/textfonts/TTRefMan/RM06/Chap6name.html +// http://www.microsoft.com/typography/otspec/name.htm + +enum { // platformID + STBTT_PLATFORM_ID_UNICODE =0, + STBTT_PLATFORM_ID_MAC =1, + STBTT_PLATFORM_ID_ISO =2, + STBTT_PLATFORM_ID_MICROSOFT =3 +}; + +enum { // encodingID for STBTT_PLATFORM_ID_UNICODE + STBTT_UNICODE_EID_UNICODE_1_0 =0, + STBTT_UNICODE_EID_UNICODE_1_1 =1, + STBTT_UNICODE_EID_ISO_10646 =2, + STBTT_UNICODE_EID_UNICODE_2_0_BMP=3, + STBTT_UNICODE_EID_UNICODE_2_0_FULL=4 +}; + +enum { // encodingID for STBTT_PLATFORM_ID_MICROSOFT + STBTT_MS_EID_SYMBOL =0, + STBTT_MS_EID_UNICODE_BMP =1, + STBTT_MS_EID_SHIFTJIS =2, + STBTT_MS_EID_UNICODE_FULL =10 +}; + +enum { // encodingID for STBTT_PLATFORM_ID_MAC; same as Script Manager codes + STBTT_MAC_EID_ROMAN =0, STBTT_MAC_EID_ARABIC =4, + STBTT_MAC_EID_JAPANESE =1, STBTT_MAC_EID_HEBREW =5, + STBTT_MAC_EID_CHINESE_TRAD =2, STBTT_MAC_EID_GREEK =6, + STBTT_MAC_EID_KOREAN =3, STBTT_MAC_EID_RUSSIAN =7 +}; + +enum { // languageID for STBTT_PLATFORM_ID_MICROSOFT; same as LCID... + // problematic because there are e.g. 16 english LCIDs and 16 arabic LCIDs + STBTT_MS_LANG_ENGLISH =0x0409, STBTT_MS_LANG_ITALIAN =0x0410, + STBTT_MS_LANG_CHINESE =0x0804, STBTT_MS_LANG_JAPANESE =0x0411, + STBTT_MS_LANG_DUTCH =0x0413, STBTT_MS_LANG_KOREAN =0x0412, + STBTT_MS_LANG_FRENCH =0x040c, STBTT_MS_LANG_RUSSIAN =0x0419, + STBTT_MS_LANG_GERMAN =0x0407, STBTT_MS_LANG_SPANISH =0x0409, + STBTT_MS_LANG_HEBREW =0x040d, STBTT_MS_LANG_SWEDISH =0x041D +}; + +enum { // languageID for STBTT_PLATFORM_ID_MAC + STBTT_MAC_LANG_ENGLISH =0 , STBTT_MAC_LANG_JAPANESE =11, + STBTT_MAC_LANG_ARABIC =12, STBTT_MAC_LANG_KOREAN =23, + STBTT_MAC_LANG_DUTCH =4 , STBTT_MAC_LANG_RUSSIAN =32, + STBTT_MAC_LANG_FRENCH =1 , STBTT_MAC_LANG_SPANISH =6 , + STBTT_MAC_LANG_GERMAN =2 , STBTT_MAC_LANG_SWEDISH =5 , + STBTT_MAC_LANG_HEBREW =10, STBTT_MAC_LANG_CHINESE_SIMPLIFIED =33, + STBTT_MAC_LANG_ITALIAN =3 , STBTT_MAC_LANG_CHINESE_TRAD =19 +}; + +#ifdef __cplusplus +} +#endif + +#endif // __STB_INCLUDE_STB_TRUETYPE_H__ + +/////////////////////////////////////////////////////////////////////////////// +/////////////////////////////////////////////////////////////////////////////// +//// +//// IMPLEMENTATION +//// +//// + +#ifdef STB_TRUETYPE_IMPLEMENTATION + +#ifndef STBTT_MAX_OVERSAMPLE +#define STBTT_MAX_OVERSAMPLE 8 +#endif + +#if STBTT_MAX_OVERSAMPLE > 255 +#error "STBTT_MAX_OVERSAMPLE cannot be > 255" +#endif + +typedef int stbtt__test_oversample_pow2[(STBTT_MAX_OVERSAMPLE & (STBTT_MAX_OVERSAMPLE-1)) == 0 ? 1 : -1]; + +#ifndef STBTT_RASTERIZER_VERSION +#define STBTT_RASTERIZER_VERSION 2 +#endif + +#ifdef _MSC_VER +#define STBTT__NOTUSED(v) (void)(v) +#else +#define STBTT__NOTUSED(v) (void)sizeof(v) +#endif + +////////////////////////////////////////////////////////////////////////// +// +// stbtt__buf helpers to parse data from file +// + +static stbtt_uint8 stbtt__buf_get8(stbtt__buf *b) +{ + if (b->cursor >= b->size) + return 0; + return b->data[b->cursor++]; +} + +static stbtt_uint8 stbtt__buf_peek8(stbtt__buf *b) +{ + if (b->cursor >= b->size) + return 0; + return b->data[b->cursor]; +} + +static void stbtt__buf_seek(stbtt__buf *b, int o) +{ + STBTT_assert(!(o > b->size || o < 0)); + b->cursor = (o > b->size || o < 0) ? b->size : o; +} + +static void stbtt__buf_skip(stbtt__buf *b, int o) +{ + stbtt__buf_seek(b, b->cursor + o); +} + +static stbtt_uint32 stbtt__buf_get(stbtt__buf *b, int n) +{ + stbtt_uint32 v = 0; + int i; + STBTT_assert(n >= 1 && n <= 4); + for (i = 0; i < n; i++) + v = (v << 8) | stbtt__buf_get8(b); + return v; +} + +static stbtt__buf stbtt__new_buf(const void *p, size_t size) +{ + stbtt__buf r; + STBTT_assert(size < 0x40000000); + r.data = (stbtt_uint8*) p; + r.size = (int) size; + r.cursor = 0; + return r; +} + +#define stbtt__buf_get16(b) stbtt__buf_get((b), 2) +#define stbtt__buf_get32(b) stbtt__buf_get((b), 4) + +static stbtt__buf stbtt__buf_range(const stbtt__buf *b, int o, int s) +{ + stbtt__buf r = stbtt__new_buf(NULL, 0); + if (o < 0 || s < 0 || o > b->size || s > b->size - o) return r; + r.data = b->data + o; + r.size = s; + return r; +} + +static stbtt__buf stbtt__cff_get_index(stbtt__buf *b) +{ + int count, start, offsize; + start = b->cursor; + count = stbtt__buf_get16(b); + if (count) { + offsize = stbtt__buf_get8(b); + STBTT_assert(offsize >= 1 && offsize <= 4); + stbtt__buf_skip(b, offsize * count); + stbtt__buf_skip(b, stbtt__buf_get(b, offsize) - 1); + } + return stbtt__buf_range(b, start, b->cursor - start); +} + +static stbtt_uint32 stbtt__cff_int(stbtt__buf *b) +{ + int b0 = stbtt__buf_get8(b); + if (b0 >= 32 && b0 <= 246) return b0 - 139; + else if (b0 >= 247 && b0 <= 250) return (b0 - 247)*256 + stbtt__buf_get8(b) + 108; + else if (b0 >= 251 && b0 <= 254) return -(b0 - 251)*256 - stbtt__buf_get8(b) - 108; + else if (b0 == 28) return stbtt__buf_get16(b); + else if (b0 == 29) return stbtt__buf_get32(b); + STBTT_assert(0); + return 0; +} + +static void stbtt__cff_skip_operand(stbtt__buf *b) { + int v, b0 = stbtt__buf_peek8(b); + STBTT_assert(b0 >= 28); + if (b0 == 30) { + stbtt__buf_skip(b, 1); + while (b->cursor < b->size) { + v = stbtt__buf_get8(b); + if ((v & 0xF) == 0xF || (v >> 4) == 0xF) + break; + } + } else { + stbtt__cff_int(b); + } +} + +static stbtt__buf stbtt__dict_get(stbtt__buf *b, int key) +{ + stbtt__buf_seek(b, 0); + while (b->cursor < b->size) { + int start = b->cursor, end, op; + while (stbtt__buf_peek8(b) >= 28) + stbtt__cff_skip_operand(b); + end = b->cursor; + op = stbtt__buf_get8(b); + if (op == 12) op = stbtt__buf_get8(b) | 0x100; + if (op == key) return stbtt__buf_range(b, start, end-start); + } + return stbtt__buf_range(b, 0, 0); +} + +static void stbtt__dict_get_ints(stbtt__buf *b, int key, int outcount, stbtt_uint32 *out) +{ + int i; + stbtt__buf operands = stbtt__dict_get(b, key); + for (i = 0; i < outcount && operands.cursor < operands.size; i++) + out[i] = stbtt__cff_int(&operands); +} + +static int stbtt__cff_index_count(stbtt__buf *b) +{ + stbtt__buf_seek(b, 0); + return stbtt__buf_get16(b); +} + +static stbtt__buf stbtt__cff_index_get(stbtt__buf b, int i) +{ + int count, offsize, start, end; + stbtt__buf_seek(&b, 0); + count = stbtt__buf_get16(&b); + offsize = stbtt__buf_get8(&b); + STBTT_assert(i >= 0 && i < count); + STBTT_assert(offsize >= 1 && offsize <= 4); + stbtt__buf_skip(&b, i*offsize); + start = stbtt__buf_get(&b, offsize); + end = stbtt__buf_get(&b, offsize); + return stbtt__buf_range(&b, 2+(count+1)*offsize+start, end - start); +} + +////////////////////////////////////////////////////////////////////////// +// +// accessors to parse data from file +// + +// on platforms that don't allow misaligned reads, if we want to allow +// truetype fonts that aren't padded to alignment, define ALLOW_UNALIGNED_TRUETYPE + +#define ttBYTE(p) (* (stbtt_uint8 *) (p)) +#define ttCHAR(p) (* (stbtt_int8 *) (p)) +#define ttFixed(p) ttLONG(p) + +static stbtt_uint16 ttUSHORT(stbtt_uint8 *p) { return p[0]*256 + p[1]; } +static stbtt_int16 ttSHORT(stbtt_uint8 *p) { return p[0]*256 + p[1]; } +static stbtt_uint32 ttULONG(stbtt_uint8 *p) { return (p[0]<<24) + (p[1]<<16) + (p[2]<<8) + p[3]; } +static stbtt_int32 ttLONG(stbtt_uint8 *p) { return (p[0]<<24) + (p[1]<<16) + (p[2]<<8) + p[3]; } + +#define stbtt_tag4(p,c0,c1,c2,c3) ((p)[0] == (c0) && (p)[1] == (c1) && (p)[2] == (c2) && (p)[3] == (c3)) +#define stbtt_tag(p,str) stbtt_tag4(p,str[0],str[1],str[2],str[3]) + +static int stbtt__isfont(stbtt_uint8 *font) +{ + // check the version number + if (stbtt_tag4(font, '1',0,0,0)) return 1; // TrueType 1 + if (stbtt_tag(font, "typ1")) return 1; // TrueType with type 1 font -- we don't support this! + if (stbtt_tag(font, "OTTO")) return 1; // OpenType with CFF + if (stbtt_tag4(font, 0,1,0,0)) return 1; // OpenType 1.0 + if (stbtt_tag(font, "true")) return 1; // Apple specification for TrueType fonts + return 0; +} + +// @OPTIMIZE: binary search +static stbtt_uint32 stbtt__find_table(stbtt_uint8 *data, stbtt_uint32 fontstart, const char *tag) +{ + stbtt_int32 num_tables = ttUSHORT(data+fontstart+4); + stbtt_uint32 tabledir = fontstart + 12; + stbtt_int32 i; + for (i=0; i < num_tables; ++i) { + stbtt_uint32 loc = tabledir + 16*i; + if (stbtt_tag(data+loc+0, tag)) + return ttULONG(data+loc+8); + } + return 0; +} + +static int stbtt_GetFontOffsetForIndex_internal(unsigned char *font_collection, int index) +{ + // if it's just a font, there's only one valid index + if (stbtt__isfont(font_collection)) + return index == 0 ? 0 : -1; + + // check if it's a TTC + if (stbtt_tag(font_collection, "ttcf")) { + // version 1? + if (ttULONG(font_collection+4) == 0x00010000 || ttULONG(font_collection+4) == 0x00020000) { + stbtt_int32 n = ttLONG(font_collection+8); + if (index >= n) + return -1; + return ttULONG(font_collection+12+index*4); + } + } + return -1; +} + +static int stbtt_GetNumberOfFonts_internal(unsigned char *font_collection) +{ + // if it's just a font, there's only one valid font + if (stbtt__isfont(font_collection)) + return 1; + + // check if it's a TTC + if (stbtt_tag(font_collection, "ttcf")) { + // version 1? + if (ttULONG(font_collection+4) == 0x00010000 || ttULONG(font_collection+4) == 0x00020000) { + return ttLONG(font_collection+8); + } + } + return 0; +} + +static stbtt__buf stbtt__get_subrs(stbtt__buf cff, stbtt__buf fontdict) +{ + stbtt_uint32 subrsoff = 0, private_loc[2] = { 0, 0 }; + stbtt__buf pdict; + stbtt__dict_get_ints(&fontdict, 18, 2, private_loc); + if (!private_loc[1] || !private_loc[0]) return stbtt__new_buf(NULL, 0); + pdict = stbtt__buf_range(&cff, private_loc[1], private_loc[0]); + stbtt__dict_get_ints(&pdict, 19, 1, &subrsoff); + if (!subrsoff) return stbtt__new_buf(NULL, 0); + stbtt__buf_seek(&cff, private_loc[1]+subrsoff); + return stbtt__cff_get_index(&cff); +} + +// since most people won't use this, find this table the first time it's needed +static int stbtt__get_svg(stbtt_fontinfo *info) +{ + stbtt_uint32 t; + if (info->svg < 0) { + t = stbtt__find_table(info->data, info->fontstart, "SVG "); + if (t) { + stbtt_uint32 offset = ttULONG(info->data + t + 2); + info->svg = t + offset; + } else { + info->svg = 0; + } + } + return info->svg; +} + +static int stbtt_InitFont_internal(stbtt_fontinfo *info, unsigned char *data, int fontstart) +{ + stbtt_uint32 cmap, t; + stbtt_int32 i,numTables; + + info->data = data; + info->fontstart = fontstart; + info->cff = stbtt__new_buf(NULL, 0); + + cmap = stbtt__find_table(data, fontstart, "cmap"); // required + info->loca = stbtt__find_table(data, fontstart, "loca"); // required + info->head = stbtt__find_table(data, fontstart, "head"); // required + info->glyf = stbtt__find_table(data, fontstart, "glyf"); // required + info->hhea = stbtt__find_table(data, fontstart, "hhea"); // required + info->hmtx = stbtt__find_table(data, fontstart, "hmtx"); // required + info->kern = stbtt__find_table(data, fontstart, "kern"); // not required + info->gpos = stbtt__find_table(data, fontstart, "GPOS"); // not required + + if (!cmap || !info->head || !info->hhea || !info->hmtx) + return 0; + if (info->glyf) { + // required for truetype + if (!info->loca) return 0; + } else { + // initialization for CFF / Type2 fonts (OTF) + stbtt__buf b, topdict, topdictidx; + stbtt_uint32 cstype = 2, charstrings = 0, fdarrayoff = 0, fdselectoff = 0; + stbtt_uint32 cff; + + cff = stbtt__find_table(data, fontstart, "CFF "); + if (!cff) return 0; + + info->fontdicts = stbtt__new_buf(NULL, 0); + info->fdselect = stbtt__new_buf(NULL, 0); + + // @TODO this should use size from table (not 512MB) + info->cff = stbtt__new_buf(data+cff, 512*1024*1024); + b = info->cff; + + // read the header + stbtt__buf_skip(&b, 2); + stbtt__buf_seek(&b, stbtt__buf_get8(&b)); // hdrsize + + // @TODO the name INDEX could list multiple fonts, + // but we just use the first one. + stbtt__cff_get_index(&b); // name INDEX + topdictidx = stbtt__cff_get_index(&b); + topdict = stbtt__cff_index_get(topdictidx, 0); + stbtt__cff_get_index(&b); // string INDEX + info->gsubrs = stbtt__cff_get_index(&b); + + stbtt__dict_get_ints(&topdict, 17, 1, &charstrings); + stbtt__dict_get_ints(&topdict, 0x100 | 6, 1, &cstype); + stbtt__dict_get_ints(&topdict, 0x100 | 36, 1, &fdarrayoff); + stbtt__dict_get_ints(&topdict, 0x100 | 37, 1, &fdselectoff); + info->subrs = stbtt__get_subrs(b, topdict); + + // we only support Type 2 charstrings + if (cstype != 2) return 0; + if (charstrings == 0) return 0; + + if (fdarrayoff) { + // looks like a CID font + if (!fdselectoff) return 0; + stbtt__buf_seek(&b, fdarrayoff); + info->fontdicts = stbtt__cff_get_index(&b); + info->fdselect = stbtt__buf_range(&b, fdselectoff, b.size-fdselectoff); + } + + stbtt__buf_seek(&b, charstrings); + info->charstrings = stbtt__cff_get_index(&b); + } + + t = stbtt__find_table(data, fontstart, "maxp"); + if (t) + info->numGlyphs = ttUSHORT(data+t+4); + else + info->numGlyphs = 0xffff; + + info->svg = -1; + + // find a cmap encoding table we understand *now* to avoid searching + // later. (todo: could make this installable) + // the same regardless of glyph. + numTables = ttUSHORT(data + cmap + 2); + info->index_map = 0; + for (i=0; i < numTables; ++i) { + stbtt_uint32 encoding_record = cmap + 4 + 8 * i; + // find an encoding we understand: + switch(ttUSHORT(data+encoding_record)) { + case STBTT_PLATFORM_ID_MICROSOFT: + switch (ttUSHORT(data+encoding_record+2)) { + case STBTT_MS_EID_UNICODE_BMP: + case STBTT_MS_EID_UNICODE_FULL: + // MS/Unicode + info->index_map = cmap + ttULONG(data+encoding_record+4); + break; + } + break; + case STBTT_PLATFORM_ID_UNICODE: + // Mac/iOS has these + // all the encodingIDs are unicode, so we don't bother to check it + info->index_map = cmap + ttULONG(data+encoding_record+4); + break; + } + } + if (info->index_map == 0) + return 0; + + info->indexToLocFormat = ttUSHORT(data+info->head + 50); + return 1; +} + +STBTT_DEF int stbtt_FindGlyphIndex(const stbtt_fontinfo *info, int unicode_codepoint) +{ + stbtt_uint8 *data = info->data; + stbtt_uint32 index_map = info->index_map; + + stbtt_uint16 format = ttUSHORT(data + index_map + 0); + if (format == 0) { // apple byte encoding + stbtt_int32 bytes = ttUSHORT(data + index_map + 2); + if (unicode_codepoint < bytes-6) + return ttBYTE(data + index_map + 6 + unicode_codepoint); + return 0; + } else if (format == 6) { + stbtt_uint32 first = ttUSHORT(data + index_map + 6); + stbtt_uint32 count = ttUSHORT(data + index_map + 8); + if ((stbtt_uint32) unicode_codepoint >= first && (stbtt_uint32) unicode_codepoint < first+count) + return ttUSHORT(data + index_map + 10 + (unicode_codepoint - first)*2); + return 0; + } else if (format == 2) { + STBTT_assert(0); // @TODO: high-byte mapping for japanese/chinese/korean + return 0; + } else if (format == 4) { // standard mapping for windows fonts: binary search collection of ranges + stbtt_uint16 segcount = ttUSHORT(data+index_map+6) >> 1; + stbtt_uint16 searchRange = ttUSHORT(data+index_map+8) >> 1; + stbtt_uint16 entrySelector = ttUSHORT(data+index_map+10); + stbtt_uint16 rangeShift = ttUSHORT(data+index_map+12) >> 1; + + // do a binary search of the segments + stbtt_uint32 endCount = index_map + 14; + stbtt_uint32 search = endCount; + + if (unicode_codepoint > 0xffff) + return 0; + + // they lie from endCount .. endCount + segCount + // but searchRange is the nearest power of two, so... + if (unicode_codepoint >= ttUSHORT(data + search + rangeShift*2)) + search += rangeShift*2; + + // now decrement to bias correctly to find smallest + search -= 2; + while (entrySelector) { + stbtt_uint16 end; + searchRange >>= 1; + end = ttUSHORT(data + search + searchRange*2); + if (unicode_codepoint > end) + search += searchRange*2; + --entrySelector; + } + search += 2; + + { + stbtt_uint16 offset, start; + stbtt_uint16 item = (stbtt_uint16) ((search - endCount) >> 1); + + STBTT_assert(unicode_codepoint <= ttUSHORT(data + endCount + 2*item)); + start = ttUSHORT(data + index_map + 14 + segcount*2 + 2 + 2*item); + if (unicode_codepoint < start) + return 0; + + offset = ttUSHORT(data + index_map + 14 + segcount*6 + 2 + 2*item); + if (offset == 0) + return (stbtt_uint16) (unicode_codepoint + ttSHORT(data + index_map + 14 + segcount*4 + 2 + 2*item)); + + return ttUSHORT(data + offset + (unicode_codepoint-start)*2 + index_map + 14 + segcount*6 + 2 + 2*item); + } + } else if (format == 12 || format == 13) { + stbtt_uint32 ngroups = ttULONG(data+index_map+12); + stbtt_int32 low,high; + low = 0; high = (stbtt_int32)ngroups; + // Binary search the right group. + while (low < high) { + stbtt_int32 mid = low + ((high-low) >> 1); // rounds down, so low <= mid < high + stbtt_uint32 start_char = ttULONG(data+index_map+16+mid*12); + stbtt_uint32 end_char = ttULONG(data+index_map+16+mid*12+4); + if ((stbtt_uint32) unicode_codepoint < start_char) + high = mid; + else if ((stbtt_uint32) unicode_codepoint > end_char) + low = mid+1; + else { + stbtt_uint32 start_glyph = ttULONG(data+index_map+16+mid*12+8); + if (format == 12) + return start_glyph + unicode_codepoint-start_char; + else // format == 13 + return start_glyph; + } + } + return 0; // not found + } + // @TODO + STBTT_assert(0); + return 0; +} + +STBTT_DEF int stbtt_GetCodepointShape(const stbtt_fontinfo *info, int unicode_codepoint, stbtt_vertex **vertices) +{ + return stbtt_GetGlyphShape(info, stbtt_FindGlyphIndex(info, unicode_codepoint), vertices); +} + +static void stbtt_setvertex(stbtt_vertex *v, stbtt_uint8 type, stbtt_int32 x, stbtt_int32 y, stbtt_int32 cx, stbtt_int32 cy) +{ + v->type = type; + v->x = (stbtt_int16) x; + v->y = (stbtt_int16) y; + v->cx = (stbtt_int16) cx; + v->cy = (stbtt_int16) cy; +} + +static int stbtt__GetGlyfOffset(const stbtt_fontinfo *info, int glyph_index) +{ + int g1,g2; + + STBTT_assert(!info->cff.size); + + if (glyph_index >= info->numGlyphs) return -1; // glyph index out of range + if (info->indexToLocFormat >= 2) return -1; // unknown index->glyph map format + + if (info->indexToLocFormat == 0) { + g1 = info->glyf + ttUSHORT(info->data + info->loca + glyph_index * 2) * 2; + g2 = info->glyf + ttUSHORT(info->data + info->loca + glyph_index * 2 + 2) * 2; + } else { + g1 = info->glyf + ttULONG (info->data + info->loca + glyph_index * 4); + g2 = info->glyf + ttULONG (info->data + info->loca + glyph_index * 4 + 4); + } + + return g1==g2 ? -1 : g1; // if length is 0, return -1 +} + +static int stbtt__GetGlyphInfoT2(const stbtt_fontinfo *info, int glyph_index, int *x0, int *y0, int *x1, int *y1); + +STBTT_DEF int stbtt_GetGlyphBox(const stbtt_fontinfo *info, int glyph_index, int *x0, int *y0, int *x1, int *y1) +{ + if (info->cff.size) { + stbtt__GetGlyphInfoT2(info, glyph_index, x0, y0, x1, y1); + } else { + int g = stbtt__GetGlyfOffset(info, glyph_index); + if (g < 0) return 0; + + if (x0) *x0 = ttSHORT(info->data + g + 2); + if (y0) *y0 = ttSHORT(info->data + g + 4); + if (x1) *x1 = ttSHORT(info->data + g + 6); + if (y1) *y1 = ttSHORT(info->data + g + 8); + } + return 1; +} + +STBTT_DEF int stbtt_GetCodepointBox(const stbtt_fontinfo *info, int codepoint, int *x0, int *y0, int *x1, int *y1) +{ + return stbtt_GetGlyphBox(info, stbtt_FindGlyphIndex(info,codepoint), x0,y0,x1,y1); +} + +STBTT_DEF int stbtt_IsGlyphEmpty(const stbtt_fontinfo *info, int glyph_index) +{ + stbtt_int16 numberOfContours; + int g; + if (info->cff.size) + return stbtt__GetGlyphInfoT2(info, glyph_index, NULL, NULL, NULL, NULL) == 0; + g = stbtt__GetGlyfOffset(info, glyph_index); + if (g < 0) return 1; + numberOfContours = ttSHORT(info->data + g); + return numberOfContours == 0; +} + +static int stbtt__close_shape(stbtt_vertex *vertices, int num_vertices, int was_off, int start_off, + stbtt_int32 sx, stbtt_int32 sy, stbtt_int32 scx, stbtt_int32 scy, stbtt_int32 cx, stbtt_int32 cy) +{ + if (start_off) { + if (was_off) + stbtt_setvertex(&vertices[num_vertices++], STBTT_vcurve, (cx+scx)>>1, (cy+scy)>>1, cx,cy); + stbtt_setvertex(&vertices[num_vertices++], STBTT_vcurve, sx,sy,scx,scy); + } else { + if (was_off) + stbtt_setvertex(&vertices[num_vertices++], STBTT_vcurve,sx,sy,cx,cy); + else + stbtt_setvertex(&vertices[num_vertices++], STBTT_vline,sx,sy,0,0); + } + return num_vertices; +} + +static int stbtt__GetGlyphShapeTT(const stbtt_fontinfo *info, int glyph_index, stbtt_vertex **pvertices) +{ + stbtt_int16 numberOfContours; + stbtt_uint8 *endPtsOfContours; + stbtt_uint8 *data = info->data; + stbtt_vertex *vertices=0; + int num_vertices=0; + int g = stbtt__GetGlyfOffset(info, glyph_index); + + *pvertices = NULL; + + if (g < 0) return 0; + + numberOfContours = ttSHORT(data + g); + + if (numberOfContours > 0) { + stbtt_uint8 flags=0,flagcount; + stbtt_int32 ins, i,j=0,m,n, next_move, was_off=0, off, start_off=0; + stbtt_int32 x,y,cx,cy,sx,sy, scx,scy; + stbtt_uint8 *points; + endPtsOfContours = (data + g + 10); + ins = ttUSHORT(data + g + 10 + numberOfContours * 2); + points = data + g + 10 + numberOfContours * 2 + 2 + ins; + + n = 1+ttUSHORT(endPtsOfContours + numberOfContours*2-2); + + m = n + 2*numberOfContours; // a loose bound on how many vertices we might need + vertices = (stbtt_vertex *) STBTT_malloc(m * sizeof(vertices[0]), info->userdata); + if (vertices == 0) + return 0; + + next_move = 0; + flagcount=0; + + // in first pass, we load uninterpreted data into the allocated array + // above, shifted to the end of the array so we won't overwrite it when + // we create our final data starting from the front + + off = m - n; // starting offset for uninterpreted data, regardless of how m ends up being calculated + + // first load flags + + for (i=0; i < n; ++i) { + if (flagcount == 0) { + flags = *points++; + if (flags & 8) + flagcount = *points++; + } else + --flagcount; + vertices[off+i].type = flags; + } + + // now load x coordinates + x=0; + for (i=0; i < n; ++i) { + flags = vertices[off+i].type; + if (flags & 2) { + stbtt_int16 dx = *points++; + x += (flags & 16) ? dx : -dx; // ??? + } else { + if (!(flags & 16)) { + x = x + (stbtt_int16) (points[0]*256 + points[1]); + points += 2; + } + } + vertices[off+i].x = (stbtt_int16) x; + } + + // now load y coordinates + y=0; + for (i=0; i < n; ++i) { + flags = vertices[off+i].type; + if (flags & 4) { + stbtt_int16 dy = *points++; + y += (flags & 32) ? dy : -dy; // ??? + } else { + if (!(flags & 32)) { + y = y + (stbtt_int16) (points[0]*256 + points[1]); + points += 2; + } + } + vertices[off+i].y = (stbtt_int16) y; + } + + // now convert them to our format + num_vertices=0; + sx = sy = cx = cy = scx = scy = 0; + for (i=0; i < n; ++i) { + flags = vertices[off+i].type; + x = (stbtt_int16) vertices[off+i].x; + y = (stbtt_int16) vertices[off+i].y; + + if (next_move == i) { + if (i != 0) + num_vertices = stbtt__close_shape(vertices, num_vertices, was_off, start_off, sx,sy,scx,scy,cx,cy); + + // now start the new one + start_off = !(flags & 1); + if (start_off) { + // if we start off with an off-curve point, then when we need to find a point on the curve + // where we can start, and we need to save some state for when we wraparound. + scx = x; + scy = y; + if (!(vertices[off+i+1].type & 1)) { + // next point is also a curve point, so interpolate an on-point curve + sx = (x + (stbtt_int32) vertices[off+i+1].x) >> 1; + sy = (y + (stbtt_int32) vertices[off+i+1].y) >> 1; + } else { + // otherwise just use the next point as our start point + sx = (stbtt_int32) vertices[off+i+1].x; + sy = (stbtt_int32) vertices[off+i+1].y; + ++i; // we're using point i+1 as the starting point, so skip it + } + } else { + sx = x; + sy = y; + } + stbtt_setvertex(&vertices[num_vertices++], STBTT_vmove,sx,sy,0,0); + was_off = 0; + next_move = 1 + ttUSHORT(endPtsOfContours+j*2); + ++j; + } else { + if (!(flags & 1)) { // if it's a curve + if (was_off) // two off-curve control points in a row means interpolate an on-curve midpoint + stbtt_setvertex(&vertices[num_vertices++], STBTT_vcurve, (cx+x)>>1, (cy+y)>>1, cx, cy); + cx = x; + cy = y; + was_off = 1; + } else { + if (was_off) + stbtt_setvertex(&vertices[num_vertices++], STBTT_vcurve, x,y, cx, cy); + else + stbtt_setvertex(&vertices[num_vertices++], STBTT_vline, x,y,0,0); + was_off = 0; + } + } + } + num_vertices = stbtt__close_shape(vertices, num_vertices, was_off, start_off, sx,sy,scx,scy,cx,cy); + } else if (numberOfContours < 0) { + // Compound shapes. + int more = 1; + stbtt_uint8 *comp = data + g + 10; + num_vertices = 0; + vertices = 0; + while (more) { + stbtt_uint16 flags, gidx; + int comp_num_verts = 0, i; + stbtt_vertex *comp_verts = 0, *tmp = 0; + float mtx[6] = {1,0,0,1,0,0}, m, n; + + flags = ttSHORT(comp); comp+=2; + gidx = ttSHORT(comp); comp+=2; + + if (flags & 2) { // XY values + if (flags & 1) { // shorts + mtx[4] = ttSHORT(comp); comp+=2; + mtx[5] = ttSHORT(comp); comp+=2; + } else { + mtx[4] = ttCHAR(comp); comp+=1; + mtx[5] = ttCHAR(comp); comp+=1; + } + } + else { + // @TODO handle matching point + STBTT_assert(0); + } + if (flags & (1<<3)) { // WE_HAVE_A_SCALE + mtx[0] = mtx[3] = ttSHORT(comp)/16384.0f; comp+=2; + mtx[1] = mtx[2] = 0; + } else if (flags & (1<<6)) { // WE_HAVE_AN_X_AND_YSCALE + mtx[0] = ttSHORT(comp)/16384.0f; comp+=2; + mtx[1] = mtx[2] = 0; + mtx[3] = ttSHORT(comp)/16384.0f; comp+=2; + } else if (flags & (1<<7)) { // WE_HAVE_A_TWO_BY_TWO + mtx[0] = ttSHORT(comp)/16384.0f; comp+=2; + mtx[1] = ttSHORT(comp)/16384.0f; comp+=2; + mtx[2] = ttSHORT(comp)/16384.0f; comp+=2; + mtx[3] = ttSHORT(comp)/16384.0f; comp+=2; + } + + // Find transformation scales. + m = (float) STBTT_sqrt(mtx[0]*mtx[0] + mtx[1]*mtx[1]); + n = (float) STBTT_sqrt(mtx[2]*mtx[2] + mtx[3]*mtx[3]); + + // Get indexed glyph. + comp_num_verts = stbtt_GetGlyphShape(info, gidx, &comp_verts); + if (comp_num_verts > 0) { + // Transform vertices. + for (i = 0; i < comp_num_verts; ++i) { + stbtt_vertex* v = &comp_verts[i]; + stbtt_vertex_type x,y; + x=v->x; y=v->y; + v->x = (stbtt_vertex_type)(m * (mtx[0]*x + mtx[2]*y + mtx[4])); + v->y = (stbtt_vertex_type)(n * (mtx[1]*x + mtx[3]*y + mtx[5])); + x=v->cx; y=v->cy; + v->cx = (stbtt_vertex_type)(m * (mtx[0]*x + mtx[2]*y + mtx[4])); + v->cy = (stbtt_vertex_type)(n * (mtx[1]*x + mtx[3]*y + mtx[5])); + } + // Append vertices. + tmp = (stbtt_vertex*)STBTT_malloc((num_vertices+comp_num_verts)*sizeof(stbtt_vertex), info->userdata); + if (!tmp) { + if (vertices) STBTT_free(vertices, info->userdata); + if (comp_verts) STBTT_free(comp_verts, info->userdata); + return 0; + } + if (num_vertices > 0) STBTT_memcpy(tmp, vertices, num_vertices*sizeof(stbtt_vertex)); + STBTT_memcpy(tmp+num_vertices, comp_verts, comp_num_verts*sizeof(stbtt_vertex)); + if (vertices) STBTT_free(vertices, info->userdata); + vertices = tmp; + STBTT_free(comp_verts, info->userdata); + num_vertices += comp_num_verts; + } + // More components ? + more = flags & (1<<5); + } + } else { + // numberOfCounters == 0, do nothing + } + + *pvertices = vertices; + return num_vertices; +} + +typedef struct +{ + int bounds; + int started; + float first_x, first_y; + float x, y; + stbtt_int32 min_x, max_x, min_y, max_y; + + stbtt_vertex *pvertices; + int num_vertices; +} stbtt__csctx; + +#define STBTT__CSCTX_INIT(bounds) {bounds,0, 0,0, 0,0, 0,0,0,0, NULL, 0} + +static void stbtt__track_vertex(stbtt__csctx *c, stbtt_int32 x, stbtt_int32 y) +{ + if (x > c->max_x || !c->started) c->max_x = x; + if (y > c->max_y || !c->started) c->max_y = y; + if (x < c->min_x || !c->started) c->min_x = x; + if (y < c->min_y || !c->started) c->min_y = y; + c->started = 1; +} + +static void stbtt__csctx_v(stbtt__csctx *c, stbtt_uint8 type, stbtt_int32 x, stbtt_int32 y, stbtt_int32 cx, stbtt_int32 cy, stbtt_int32 cx1, stbtt_int32 cy1) +{ + if (c->bounds) { + stbtt__track_vertex(c, x, y); + if (type == STBTT_vcubic) { + stbtt__track_vertex(c, cx, cy); + stbtt__track_vertex(c, cx1, cy1); + } + } else { + stbtt_setvertex(&c->pvertices[c->num_vertices], type, x, y, cx, cy); + c->pvertices[c->num_vertices].cx1 = (stbtt_int16) cx1; + c->pvertices[c->num_vertices].cy1 = (stbtt_int16) cy1; + } + c->num_vertices++; +} + +static void stbtt__csctx_close_shape(stbtt__csctx *ctx) +{ + if (ctx->first_x != ctx->x || ctx->first_y != ctx->y) + stbtt__csctx_v(ctx, STBTT_vline, (int)ctx->first_x, (int)ctx->first_y, 0, 0, 0, 0); +} + +static void stbtt__csctx_rmove_to(stbtt__csctx *ctx, float dx, float dy) +{ + stbtt__csctx_close_shape(ctx); + ctx->first_x = ctx->x = ctx->x + dx; + ctx->first_y = ctx->y = ctx->y + dy; + stbtt__csctx_v(ctx, STBTT_vmove, (int)ctx->x, (int)ctx->y, 0, 0, 0, 0); +} + +static void stbtt__csctx_rline_to(stbtt__csctx *ctx, float dx, float dy) +{ + ctx->x += dx; + ctx->y += dy; + stbtt__csctx_v(ctx, STBTT_vline, (int)ctx->x, (int)ctx->y, 0, 0, 0, 0); +} + +static void stbtt__csctx_rccurve_to(stbtt__csctx *ctx, float dx1, float dy1, float dx2, float dy2, float dx3, float dy3) +{ + float cx1 = ctx->x + dx1; + float cy1 = ctx->y + dy1; + float cx2 = cx1 + dx2; + float cy2 = cy1 + dy2; + ctx->x = cx2 + dx3; + ctx->y = cy2 + dy3; + stbtt__csctx_v(ctx, STBTT_vcubic, (int)ctx->x, (int)ctx->y, (int)cx1, (int)cy1, (int)cx2, (int)cy2); +} + +static stbtt__buf stbtt__get_subr(stbtt__buf idx, int n) +{ + int count = stbtt__cff_index_count(&idx); + int bias = 107; + if (count >= 33900) + bias = 32768; + else if (count >= 1240) + bias = 1131; + n += bias; + if (n < 0 || n >= count) + return stbtt__new_buf(NULL, 0); + return stbtt__cff_index_get(idx, n); +} + +static stbtt__buf stbtt__cid_get_glyph_subrs(const stbtt_fontinfo *info, int glyph_index) +{ + stbtt__buf fdselect = info->fdselect; + int nranges, start, end, v, fmt, fdselector = -1, i; + + stbtt__buf_seek(&fdselect, 0); + fmt = stbtt__buf_get8(&fdselect); + if (fmt == 0) { + // untested + stbtt__buf_skip(&fdselect, glyph_index); + fdselector = stbtt__buf_get8(&fdselect); + } else if (fmt == 3) { + nranges = stbtt__buf_get16(&fdselect); + start = stbtt__buf_get16(&fdselect); + for (i = 0; i < nranges; i++) { + v = stbtt__buf_get8(&fdselect); + end = stbtt__buf_get16(&fdselect); + if (glyph_index >= start && glyph_index < end) { + fdselector = v; + break; + } + start = end; + } + } + if (fdselector == -1) stbtt__new_buf(NULL, 0); + return stbtt__get_subrs(info->cff, stbtt__cff_index_get(info->fontdicts, fdselector)); +} + +static int stbtt__run_charstring(const stbtt_fontinfo *info, int glyph_index, stbtt__csctx *c) +{ + int in_header = 1, maskbits = 0, subr_stack_height = 0, sp = 0, v, i, b0; + int has_subrs = 0, clear_stack; + float s[48]; + stbtt__buf subr_stack[10], subrs = info->subrs, b; + float f; + +#define STBTT__CSERR(s) (0) + + // this currently ignores the initial width value, which isn't needed if we have hmtx + b = stbtt__cff_index_get(info->charstrings, glyph_index); + while (b.cursor < b.size) { + i = 0; + clear_stack = 1; + b0 = stbtt__buf_get8(&b); + switch (b0) { + // @TODO implement hinting + case 0x13: // hintmask + case 0x14: // cntrmask + if (in_header) + maskbits += (sp / 2); // implicit "vstem" + in_header = 0; + stbtt__buf_skip(&b, (maskbits + 7) / 8); + break; + + case 0x01: // hstem + case 0x03: // vstem + case 0x12: // hstemhm + case 0x17: // vstemhm + maskbits += (sp / 2); + break; + + case 0x15: // rmoveto + in_header = 0; + if (sp < 2) return STBTT__CSERR("rmoveto stack"); + stbtt__csctx_rmove_to(c, s[sp-2], s[sp-1]); + break; + case 0x04: // vmoveto + in_header = 0; + if (sp < 1) return STBTT__CSERR("vmoveto stack"); + stbtt__csctx_rmove_to(c, 0, s[sp-1]); + break; + case 0x16: // hmoveto + in_header = 0; + if (sp < 1) return STBTT__CSERR("hmoveto stack"); + stbtt__csctx_rmove_to(c, s[sp-1], 0); + break; + + case 0x05: // rlineto + if (sp < 2) return STBTT__CSERR("rlineto stack"); + for (; i + 1 < sp; i += 2) + stbtt__csctx_rline_to(c, s[i], s[i+1]); + break; + + // hlineto/vlineto and vhcurveto/hvcurveto alternate horizontal and vertical + // starting from a different place. + + case 0x07: // vlineto + if (sp < 1) return STBTT__CSERR("vlineto stack"); + goto vlineto; + case 0x06: // hlineto + if (sp < 1) return STBTT__CSERR("hlineto stack"); + for (;;) { + if (i >= sp) break; + stbtt__csctx_rline_to(c, s[i], 0); + i++; + vlineto: + if (i >= sp) break; + stbtt__csctx_rline_to(c, 0, s[i]); + i++; + } + break; + + case 0x1F: // hvcurveto + if (sp < 4) return STBTT__CSERR("hvcurveto stack"); + goto hvcurveto; + case 0x1E: // vhcurveto + if (sp < 4) return STBTT__CSERR("vhcurveto stack"); + for (;;) { + if (i + 3 >= sp) break; + stbtt__csctx_rccurve_to(c, 0, s[i], s[i+1], s[i+2], s[i+3], (sp - i == 5) ? s[i + 4] : 0.0f); + i += 4; + hvcurveto: + if (i + 3 >= sp) break; + stbtt__csctx_rccurve_to(c, s[i], 0, s[i+1], s[i+2], (sp - i == 5) ? s[i+4] : 0.0f, s[i+3]); + i += 4; + } + break; + + case 0x08: // rrcurveto + if (sp < 6) return STBTT__CSERR("rcurveline stack"); + for (; i + 5 < sp; i += 6) + stbtt__csctx_rccurve_to(c, s[i], s[i+1], s[i+2], s[i+3], s[i+4], s[i+5]); + break; + + case 0x18: // rcurveline + if (sp < 8) return STBTT__CSERR("rcurveline stack"); + for (; i + 5 < sp - 2; i += 6) + stbtt__csctx_rccurve_to(c, s[i], s[i+1], s[i+2], s[i+3], s[i+4], s[i+5]); + if (i + 1 >= sp) return STBTT__CSERR("rcurveline stack"); + stbtt__csctx_rline_to(c, s[i], s[i+1]); + break; + + case 0x19: // rlinecurve + if (sp < 8) return STBTT__CSERR("rlinecurve stack"); + for (; i + 1 < sp - 6; i += 2) + stbtt__csctx_rline_to(c, s[i], s[i+1]); + if (i + 5 >= sp) return STBTT__CSERR("rlinecurve stack"); + stbtt__csctx_rccurve_to(c, s[i], s[i+1], s[i+2], s[i+3], s[i+4], s[i+5]); + break; + + case 0x1A: // vvcurveto + case 0x1B: // hhcurveto + if (sp < 4) return STBTT__CSERR("(vv|hh)curveto stack"); + f = 0.0; + if (sp & 1) { f = s[i]; i++; } + for (; i + 3 < sp; i += 4) { + if (b0 == 0x1B) + stbtt__csctx_rccurve_to(c, s[i], f, s[i+1], s[i+2], s[i+3], 0.0); + else + stbtt__csctx_rccurve_to(c, f, s[i], s[i+1], s[i+2], 0.0, s[i+3]); + f = 0.0; + } + break; + + case 0x0A: // callsubr + if (!has_subrs) { + if (info->fdselect.size) + subrs = stbtt__cid_get_glyph_subrs(info, glyph_index); + has_subrs = 1; + } + // fallthrough + case 0x1D: // callgsubr + if (sp < 1) return STBTT__CSERR("call(g|)subr stack"); + v = (int) s[--sp]; + if (subr_stack_height >= 10) return STBTT__CSERR("recursion limit"); + subr_stack[subr_stack_height++] = b; + b = stbtt__get_subr(b0 == 0x0A ? subrs : info->gsubrs, v); + if (b.size == 0) return STBTT__CSERR("subr not found"); + b.cursor = 0; + clear_stack = 0; + break; + + case 0x0B: // return + if (subr_stack_height <= 0) return STBTT__CSERR("return outside subr"); + b = subr_stack[--subr_stack_height]; + clear_stack = 0; + break; + + case 0x0E: // endchar + stbtt__csctx_close_shape(c); + return 1; + + case 0x0C: { // two-byte escape + float dx1, dx2, dx3, dx4, dx5, dx6, dy1, dy2, dy3, dy4, dy5, dy6; + float dx, dy; + int b1 = stbtt__buf_get8(&b); + switch (b1) { + // @TODO These "flex" implementations ignore the flex-depth and resolution, + // and always draw beziers. + case 0x22: // hflex + if (sp < 7) return STBTT__CSERR("hflex stack"); + dx1 = s[0]; + dx2 = s[1]; + dy2 = s[2]; + dx3 = s[3]; + dx4 = s[4]; + dx5 = s[5]; + dx6 = s[6]; + stbtt__csctx_rccurve_to(c, dx1, 0, dx2, dy2, dx3, 0); + stbtt__csctx_rccurve_to(c, dx4, 0, dx5, -dy2, dx6, 0); + break; + + case 0x23: // flex + if (sp < 13) return STBTT__CSERR("flex stack"); + dx1 = s[0]; + dy1 = s[1]; + dx2 = s[2]; + dy2 = s[3]; + dx3 = s[4]; + dy3 = s[5]; + dx4 = s[6]; + dy4 = s[7]; + dx5 = s[8]; + dy5 = s[9]; + dx6 = s[10]; + dy6 = s[11]; + //fd is s[12] + stbtt__csctx_rccurve_to(c, dx1, dy1, dx2, dy2, dx3, dy3); + stbtt__csctx_rccurve_to(c, dx4, dy4, dx5, dy5, dx6, dy6); + break; + + case 0x24: // hflex1 + if (sp < 9) return STBTT__CSERR("hflex1 stack"); + dx1 = s[0]; + dy1 = s[1]; + dx2 = s[2]; + dy2 = s[3]; + dx3 = s[4]; + dx4 = s[5]; + dx5 = s[6]; + dy5 = s[7]; + dx6 = s[8]; + stbtt__csctx_rccurve_to(c, dx1, dy1, dx2, dy2, dx3, 0); + stbtt__csctx_rccurve_to(c, dx4, 0, dx5, dy5, dx6, -(dy1+dy2+dy5)); + break; + + case 0x25: // flex1 + if (sp < 11) return STBTT__CSERR("flex1 stack"); + dx1 = s[0]; + dy1 = s[1]; + dx2 = s[2]; + dy2 = s[3]; + dx3 = s[4]; + dy3 = s[5]; + dx4 = s[6]; + dy4 = s[7]; + dx5 = s[8]; + dy5 = s[9]; + dx6 = dy6 = s[10]; + dx = dx1+dx2+dx3+dx4+dx5; + dy = dy1+dy2+dy3+dy4+dy5; + if (STBTT_fabs(dx) > STBTT_fabs(dy)) + dy6 = -dy; + else + dx6 = -dx; + stbtt__csctx_rccurve_to(c, dx1, dy1, dx2, dy2, dx3, dy3); + stbtt__csctx_rccurve_to(c, dx4, dy4, dx5, dy5, dx6, dy6); + break; + + default: + return STBTT__CSERR("unimplemented"); + } + } break; + + default: + if (b0 != 255 && b0 != 28 && (b0 < 32 || b0 > 254)) + return STBTT__CSERR("reserved operator"); + + // push immediate + if (b0 == 255) { + f = (float)(stbtt_int32)stbtt__buf_get32(&b) / 0x10000; + } else { + stbtt__buf_skip(&b, -1); + f = (float)(stbtt_int16)stbtt__cff_int(&b); + } + if (sp >= 48) return STBTT__CSERR("push stack overflow"); + s[sp++] = f; + clear_stack = 0; + break; + } + if (clear_stack) sp = 0; + } + return STBTT__CSERR("no endchar"); + +#undef STBTT__CSERR +} + +static int stbtt__GetGlyphShapeT2(const stbtt_fontinfo *info, int glyph_index, stbtt_vertex **pvertices) +{ + // runs the charstring twice, once to count and once to output (to avoid realloc) + stbtt__csctx count_ctx = STBTT__CSCTX_INIT(1); + stbtt__csctx output_ctx = STBTT__CSCTX_INIT(0); + if (stbtt__run_charstring(info, glyph_index, &count_ctx)) { + *pvertices = (stbtt_vertex*)STBTT_malloc(count_ctx.num_vertices*sizeof(stbtt_vertex), info->userdata); + output_ctx.pvertices = *pvertices; + if (stbtt__run_charstring(info, glyph_index, &output_ctx)) { + STBTT_assert(output_ctx.num_vertices == count_ctx.num_vertices); + return output_ctx.num_vertices; + } + } + *pvertices = NULL; + return 0; +} + +static int stbtt__GetGlyphInfoT2(const stbtt_fontinfo *info, int glyph_index, int *x0, int *y0, int *x1, int *y1) +{ + stbtt__csctx c = STBTT__CSCTX_INIT(1); + int r = stbtt__run_charstring(info, glyph_index, &c); + if (x0) *x0 = r ? c.min_x : 0; + if (y0) *y0 = r ? c.min_y : 0; + if (x1) *x1 = r ? c.max_x : 0; + if (y1) *y1 = r ? c.max_y : 0; + return r ? c.num_vertices : 0; +} + +STBTT_DEF int stbtt_GetGlyphShape(const stbtt_fontinfo *info, int glyph_index, stbtt_vertex **pvertices) +{ + if (!info->cff.size) + return stbtt__GetGlyphShapeTT(info, glyph_index, pvertices); + else + return stbtt__GetGlyphShapeT2(info, glyph_index, pvertices); +} + +STBTT_DEF void stbtt_GetGlyphHMetrics(const stbtt_fontinfo *info, int glyph_index, int *advanceWidth, int *leftSideBearing) +{ + stbtt_uint16 numOfLongHorMetrics = ttUSHORT(info->data+info->hhea + 34); + if (glyph_index < numOfLongHorMetrics) { + if (advanceWidth) *advanceWidth = ttSHORT(info->data + info->hmtx + 4*glyph_index); + if (leftSideBearing) *leftSideBearing = ttSHORT(info->data + info->hmtx + 4*glyph_index + 2); + } else { + if (advanceWidth) *advanceWidth = ttSHORT(info->data + info->hmtx + 4*(numOfLongHorMetrics-1)); + if (leftSideBearing) *leftSideBearing = ttSHORT(info->data + info->hmtx + 4*numOfLongHorMetrics + 2*(glyph_index - numOfLongHorMetrics)); + } +} + +STBTT_DEF int stbtt_GetKerningTableLength(const stbtt_fontinfo *info) +{ + stbtt_uint8 *data = info->data + info->kern; + + // we only look at the first table. it must be 'horizontal' and format 0. + if (!info->kern) + return 0; + if (ttUSHORT(data+2) < 1) // number of tables, need at least 1 + return 0; + if (ttUSHORT(data+8) != 1) // horizontal flag must be set in format + return 0; + + return ttUSHORT(data+10); +} + +STBTT_DEF int stbtt_GetKerningTable(const stbtt_fontinfo *info, stbtt_kerningentry* table, int table_length) +{ + stbtt_uint8 *data = info->data + info->kern; + int k, length; + + // we only look at the first table. it must be 'horizontal' and format 0. + if (!info->kern) + return 0; + if (ttUSHORT(data+2) < 1) // number of tables, need at least 1 + return 0; + if (ttUSHORT(data+8) != 1) // horizontal flag must be set in format + return 0; + + length = ttUSHORT(data+10); + if (table_length < length) + length = table_length; + + for (k = 0; k < length; k++) + { + table[k].glyph1 = ttUSHORT(data+18+(k*6)); + table[k].glyph2 = ttUSHORT(data+20+(k*6)); + table[k].advance = ttSHORT(data+22+(k*6)); + } + + return length; +} + +static int stbtt__GetGlyphKernInfoAdvance(const stbtt_fontinfo *info, int glyph1, int glyph2) +{ + stbtt_uint8 *data = info->data + info->kern; + stbtt_uint32 needle, straw; + int l, r, m; + + // we only look at the first table. it must be 'horizontal' and format 0. + if (!info->kern) + return 0; + if (ttUSHORT(data+2) < 1) // number of tables, need at least 1 + return 0; + if (ttUSHORT(data+8) != 1) // horizontal flag must be set in format + return 0; + + l = 0; + r = ttUSHORT(data+10) - 1; + needle = glyph1 << 16 | glyph2; + while (l <= r) { + m = (l + r) >> 1; + straw = ttULONG(data+18+(m*6)); // note: unaligned read + if (needle < straw) + r = m - 1; + else if (needle > straw) + l = m + 1; + else + return ttSHORT(data+22+(m*6)); + } + return 0; +} + +static stbtt_int32 stbtt__GetCoverageIndex(stbtt_uint8 *coverageTable, int glyph) +{ + stbtt_uint16 coverageFormat = ttUSHORT(coverageTable); + switch(coverageFormat) { + case 1: { + stbtt_uint16 glyphCount = ttUSHORT(coverageTable + 2); + + // Binary search. + stbtt_int32 l=0, r=glyphCount-1, m; + int straw, needle=glyph; + while (l <= r) { + stbtt_uint8 *glyphArray = coverageTable + 4; + stbtt_uint16 glyphID; + m = (l + r) >> 1; + glyphID = ttUSHORT(glyphArray + 2 * m); + straw = glyphID; + if (needle < straw) + r = m - 1; + else if (needle > straw) + l = m + 1; + else { + return m; + } + } + } break; + + case 2: { + stbtt_uint16 rangeCount = ttUSHORT(coverageTable + 2); + stbtt_uint8 *rangeArray = coverageTable + 4; + + // Binary search. + stbtt_int32 l=0, r=rangeCount-1, m; + int strawStart, strawEnd, needle=glyph; + while (l <= r) { + stbtt_uint8 *rangeRecord; + m = (l + r) >> 1; + rangeRecord = rangeArray + 6 * m; + strawStart = ttUSHORT(rangeRecord); + strawEnd = ttUSHORT(rangeRecord + 2); + if (needle < strawStart) + r = m - 1; + else if (needle > strawEnd) + l = m + 1; + else { + stbtt_uint16 startCoverageIndex = ttUSHORT(rangeRecord + 4); + return startCoverageIndex + glyph - strawStart; + } + } + } break; + + default: { + // There are no other cases. + STBTT_assert(0); + } break; + } + + return -1; +} + +static stbtt_int32 stbtt__GetGlyphClass(stbtt_uint8 *classDefTable, int glyph) +{ + stbtt_uint16 classDefFormat = ttUSHORT(classDefTable); + switch(classDefFormat) + { + case 1: { + stbtt_uint16 startGlyphID = ttUSHORT(classDefTable + 2); + stbtt_uint16 glyphCount = ttUSHORT(classDefTable + 4); + stbtt_uint8 *classDef1ValueArray = classDefTable + 6; + + if (glyph >= startGlyphID && glyph < startGlyphID + glyphCount) + return (stbtt_int32)ttUSHORT(classDef1ValueArray + 2 * (glyph - startGlyphID)); + + classDefTable = classDef1ValueArray + 2 * glyphCount; + } break; + + case 2: { + stbtt_uint16 classRangeCount = ttUSHORT(classDefTable + 2); + stbtt_uint8 *classRangeRecords = classDefTable + 4; + + // Binary search. + stbtt_int32 l=0, r=classRangeCount-1, m; + int strawStart, strawEnd, needle=glyph; + while (l <= r) { + stbtt_uint8 *classRangeRecord; + m = (l + r) >> 1; + classRangeRecord = classRangeRecords + 6 * m; + strawStart = ttUSHORT(classRangeRecord); + strawEnd = ttUSHORT(classRangeRecord + 2); + if (needle < strawStart) + r = m - 1; + else if (needle > strawEnd) + l = m + 1; + else + return (stbtt_int32)ttUSHORT(classRangeRecord + 4); + } + + classDefTable = classRangeRecords + 6 * classRangeCount; + } break; + + default: { + // There are no other cases. + STBTT_assert(0); + } break; + } + + return -1; +} + +// Define to STBTT_assert(x) if you want to break on unimplemented formats. +#define STBTT_GPOS_TODO_assert(x) + +static stbtt_int32 stbtt__GetGlyphGPOSInfoAdvance(const stbtt_fontinfo *info, int glyph1, int glyph2) +{ + stbtt_uint16 lookupListOffset; + stbtt_uint8 *lookupList; + stbtt_uint16 lookupCount; + stbtt_uint8 *data; + stbtt_int32 i; + + if (!info->gpos) return 0; + + data = info->data + info->gpos; + + if (ttUSHORT(data+0) != 1) return 0; // Major version 1 + if (ttUSHORT(data+2) != 0) return 0; // Minor version 0 + + lookupListOffset = ttUSHORT(data+8); + lookupList = data + lookupListOffset; + lookupCount = ttUSHORT(lookupList); + + for (i=0; i> 1; + pairValue = pairValueArray + (2 + valueRecordPairSizeInBytes) * m; + secondGlyph = ttUSHORT(pairValue); + straw = secondGlyph; + if (needle < straw) + r = m - 1; + else if (needle > straw) + l = m + 1; + else { + stbtt_int16 xAdvance = ttSHORT(pairValue + 2); + return xAdvance; + } + } + } break; + + case 2: { + stbtt_uint16 valueFormat1 = ttUSHORT(table + 4); + stbtt_uint16 valueFormat2 = ttUSHORT(table + 6); + + stbtt_uint16 classDef1Offset = ttUSHORT(table + 8); + stbtt_uint16 classDef2Offset = ttUSHORT(table + 10); + int glyph1class = stbtt__GetGlyphClass(table + classDef1Offset, glyph1); + int glyph2class = stbtt__GetGlyphClass(table + classDef2Offset, glyph2); + + stbtt_uint16 class1Count = ttUSHORT(table + 12); + stbtt_uint16 class2Count = ttUSHORT(table + 14); + STBTT_assert(glyph1class < class1Count); + STBTT_assert(glyph2class < class2Count); + + // TODO: Support more formats. + STBTT_GPOS_TODO_assert(valueFormat1 == 4); + if (valueFormat1 != 4) return 0; + STBTT_GPOS_TODO_assert(valueFormat2 == 0); + if (valueFormat2 != 0) return 0; + + if (glyph1class >= 0 && glyph1class < class1Count && glyph2class >= 0 && glyph2class < class2Count) { + stbtt_uint8 *class1Records = table + 16; + stbtt_uint8 *class2Records = class1Records + 2 * (glyph1class * class2Count); + stbtt_int16 xAdvance = ttSHORT(class2Records + 2 * glyph2class); + return xAdvance; + } + } break; + + default: { + // There are no other cases. + STBTT_assert(0); + break; + }; + } + } + break; + }; + + default: + // TODO: Implement other stuff. + break; + } + } + + return 0; +} + +STBTT_DEF int stbtt_GetGlyphKernAdvance(const stbtt_fontinfo *info, int g1, int g2) +{ + int xAdvance = 0; + + if (info->gpos) + xAdvance += stbtt__GetGlyphGPOSInfoAdvance(info, g1, g2); + else if (info->kern) + xAdvance += stbtt__GetGlyphKernInfoAdvance(info, g1, g2); + + return xAdvance; +} + +STBTT_DEF int stbtt_GetCodepointKernAdvance(const stbtt_fontinfo *info, int ch1, int ch2) +{ + if (!info->kern && !info->gpos) // if no kerning table, don't waste time looking up both codepoint->glyphs + return 0; + return stbtt_GetGlyphKernAdvance(info, stbtt_FindGlyphIndex(info,ch1), stbtt_FindGlyphIndex(info,ch2)); +} + +STBTT_DEF void stbtt_GetCodepointHMetrics(const stbtt_fontinfo *info, int codepoint, int *advanceWidth, int *leftSideBearing) +{ + stbtt_GetGlyphHMetrics(info, stbtt_FindGlyphIndex(info,codepoint), advanceWidth, leftSideBearing); +} + +STBTT_DEF void stbtt_GetFontVMetrics(const stbtt_fontinfo *info, int *ascent, int *descent, int *lineGap) +{ + if (ascent ) *ascent = ttSHORT(info->data+info->hhea + 4); + if (descent) *descent = ttSHORT(info->data+info->hhea + 6); + if (lineGap) *lineGap = ttSHORT(info->data+info->hhea + 8); +} + +STBTT_DEF int stbtt_GetFontVMetricsOS2(const stbtt_fontinfo *info, int *typoAscent, int *typoDescent, int *typoLineGap) +{ + int tab = stbtt__find_table(info->data, info->fontstart, "OS/2"); + if (!tab) + return 0; + if (typoAscent ) *typoAscent = ttSHORT(info->data+tab + 68); + if (typoDescent) *typoDescent = ttSHORT(info->data+tab + 70); + if (typoLineGap) *typoLineGap = ttSHORT(info->data+tab + 72); + return 1; +} + +STBTT_DEF void stbtt_GetFontBoundingBox(const stbtt_fontinfo *info, int *x0, int *y0, int *x1, int *y1) +{ + *x0 = ttSHORT(info->data + info->head + 36); + *y0 = ttSHORT(info->data + info->head + 38); + *x1 = ttSHORT(info->data + info->head + 40); + *y1 = ttSHORT(info->data + info->head + 42); +} + +STBTT_DEF float stbtt_ScaleForPixelHeight(const stbtt_fontinfo *info, float height) +{ + int fheight = ttSHORT(info->data + info->hhea + 4) - ttSHORT(info->data + info->hhea + 6); + return (float) height / fheight; +} + +STBTT_DEF float stbtt_ScaleForMappingEmToPixels(const stbtt_fontinfo *info, float pixels) +{ + int unitsPerEm = ttUSHORT(info->data + info->head + 18); + return pixels / unitsPerEm; +} + +STBTT_DEF void stbtt_FreeShape(const stbtt_fontinfo *info, stbtt_vertex *v) +{ + STBTT_free(v, info->userdata); +} + +STBTT_DEF stbtt_uint8 *stbtt_FindSVGDoc(const stbtt_fontinfo *info, int gl) +{ + int i; + stbtt_uint8 *data = info->data; + stbtt_uint8 *svg_doc_list = data + stbtt__get_svg((stbtt_fontinfo *) info); + + int numEntries = ttUSHORT(svg_doc_list); + stbtt_uint8 *svg_docs = svg_doc_list + 2; + + for(i=0; i= ttUSHORT(svg_doc)) && (gl <= ttUSHORT(svg_doc + 2))) + return svg_doc; + } + return 0; +} + +STBTT_DEF int stbtt_GetGlyphSVG(const stbtt_fontinfo *info, int gl, const char **svg) +{ + stbtt_uint8 *data = info->data; + stbtt_uint8 *svg_doc; + + if (info->svg == 0) + return 0; + + svg_doc = stbtt_FindSVGDoc(info, gl); + if (svg_doc != NULL) { + *svg = (char *) data + info->svg + ttULONG(svg_doc + 4); + return ttULONG(svg_doc + 8); + } else { + return 0; + } +} + +STBTT_DEF int stbtt_GetCodepointSVG(const stbtt_fontinfo *info, int unicode_codepoint, const char **svg) +{ + return stbtt_GetGlyphSVG(info, stbtt_FindGlyphIndex(info, unicode_codepoint), svg); +} + +////////////////////////////////////////////////////////////////////////////// +// +// antialiasing software rasterizer +// + +STBTT_DEF void stbtt_GetGlyphBitmapBoxSubpixel(const stbtt_fontinfo *font, int glyph, float scale_x, float scale_y,float shift_x, float shift_y, int *ix0, int *iy0, int *ix1, int *iy1) +{ + int x0=0,y0=0,x1,y1; // =0 suppresses compiler warning + if (!stbtt_GetGlyphBox(font, glyph, &x0,&y0,&x1,&y1)) { + // e.g. space character + if (ix0) *ix0 = 0; + if (iy0) *iy0 = 0; + if (ix1) *ix1 = 0; + if (iy1) *iy1 = 0; + } else { + // move to integral bboxes (treating pixels as little squares, what pixels get touched)? + if (ix0) *ix0 = STBTT_ifloor( x0 * scale_x + shift_x); + if (iy0) *iy0 = STBTT_ifloor(-y1 * scale_y + shift_y); + if (ix1) *ix1 = STBTT_iceil ( x1 * scale_x + shift_x); + if (iy1) *iy1 = STBTT_iceil (-y0 * scale_y + shift_y); + } +} + +STBTT_DEF void stbtt_GetGlyphBitmapBox(const stbtt_fontinfo *font, int glyph, float scale_x, float scale_y, int *ix0, int *iy0, int *ix1, int *iy1) +{ + stbtt_GetGlyphBitmapBoxSubpixel(font, glyph, scale_x, scale_y,0.0f,0.0f, ix0, iy0, ix1, iy1); +} + +STBTT_DEF void stbtt_GetCodepointBitmapBoxSubpixel(const stbtt_fontinfo *font, int codepoint, float scale_x, float scale_y, float shift_x, float shift_y, int *ix0, int *iy0, int *ix1, int *iy1) +{ + stbtt_GetGlyphBitmapBoxSubpixel(font, stbtt_FindGlyphIndex(font,codepoint), scale_x, scale_y,shift_x,shift_y, ix0,iy0,ix1,iy1); +} + +STBTT_DEF void stbtt_GetCodepointBitmapBox(const stbtt_fontinfo *font, int codepoint, float scale_x, float scale_y, int *ix0, int *iy0, int *ix1, int *iy1) +{ + stbtt_GetCodepointBitmapBoxSubpixel(font, codepoint, scale_x, scale_y,0.0f,0.0f, ix0,iy0,ix1,iy1); +} + +////////////////////////////////////////////////////////////////////////////// +// +// Rasterizer + +typedef struct stbtt__hheap_chunk +{ + struct stbtt__hheap_chunk *next; +} stbtt__hheap_chunk; + +typedef struct stbtt__hheap +{ + struct stbtt__hheap_chunk *head; + void *first_free; + int num_remaining_in_head_chunk; +} stbtt__hheap; + +static void *stbtt__hheap_alloc(stbtt__hheap *hh, size_t size, void *userdata) +{ + if (hh->first_free) { + void *p = hh->first_free; + hh->first_free = * (void **) p; + return p; + } else { + if (hh->num_remaining_in_head_chunk == 0) { + int count = (size < 32 ? 2000 : size < 128 ? 800 : 100); + stbtt__hheap_chunk *c = (stbtt__hheap_chunk *) STBTT_malloc(sizeof(stbtt__hheap_chunk) + size * count, userdata); + if (c == NULL) + return NULL; + c->next = hh->head; + hh->head = c; + hh->num_remaining_in_head_chunk = count; + } + --hh->num_remaining_in_head_chunk; + return (char *) (hh->head) + sizeof(stbtt__hheap_chunk) + size * hh->num_remaining_in_head_chunk; + } +} + +static void stbtt__hheap_free(stbtt__hheap *hh, void *p) +{ + *(void **) p = hh->first_free; + hh->first_free = p; +} + +static void stbtt__hheap_cleanup(stbtt__hheap *hh, void *userdata) +{ + stbtt__hheap_chunk *c = hh->head; + while (c) { + stbtt__hheap_chunk *n = c->next; + STBTT_free(c, userdata); + c = n; + } +} + +typedef struct stbtt__edge { + float x0,y0, x1,y1; + int invert; +} stbtt__edge; + + +typedef struct stbtt__active_edge +{ + struct stbtt__active_edge *next; + #if STBTT_RASTERIZER_VERSION==1 + int x,dx; + float ey; + int direction; + #elif STBTT_RASTERIZER_VERSION==2 + float fx,fdx,fdy; + float direction; + float sy; + float ey; + #else + #error "Unrecognized value of STBTT_RASTERIZER_VERSION" + #endif +} stbtt__active_edge; + +#if STBTT_RASTERIZER_VERSION == 1 +#define STBTT_FIXSHIFT 10 +#define STBTT_FIX (1 << STBTT_FIXSHIFT) +#define STBTT_FIXMASK (STBTT_FIX-1) + +static stbtt__active_edge *stbtt__new_active(stbtt__hheap *hh, stbtt__edge *e, int off_x, float start_point, void *userdata) +{ + stbtt__active_edge *z = (stbtt__active_edge *) stbtt__hheap_alloc(hh, sizeof(*z), userdata); + float dxdy = (e->x1 - e->x0) / (e->y1 - e->y0); + STBTT_assert(z != NULL); + if (!z) return z; + + // round dx down to avoid overshooting + if (dxdy < 0) + z->dx = -STBTT_ifloor(STBTT_FIX * -dxdy); + else + z->dx = STBTT_ifloor(STBTT_FIX * dxdy); + + z->x = STBTT_ifloor(STBTT_FIX * e->x0 + z->dx * (start_point - e->y0)); // use z->dx so when we offset later it's by the same amount + z->x -= off_x * STBTT_FIX; + + z->ey = e->y1; + z->next = 0; + z->direction = e->invert ? 1 : -1; + return z; +} +#elif STBTT_RASTERIZER_VERSION == 2 +static stbtt__active_edge *stbtt__new_active(stbtt__hheap *hh, stbtt__edge *e, int off_x, float start_point, void *userdata) +{ + stbtt__active_edge *z = (stbtt__active_edge *) stbtt__hheap_alloc(hh, sizeof(*z), userdata); + float dxdy = (e->x1 - e->x0) / (e->y1 - e->y0); + STBTT_assert(z != NULL); + //STBTT_assert(e->y0 <= start_point); + if (!z) return z; + z->fdx = dxdy; + z->fdy = dxdy != 0.0f ? (1.0f/dxdy) : 0.0f; + z->fx = e->x0 + dxdy * (start_point - e->y0); + z->fx -= off_x; + z->direction = e->invert ? 1.0f : -1.0f; + z->sy = e->y0; + z->ey = e->y1; + z->next = 0; + return z; +} +#else +#error "Unrecognized value of STBTT_RASTERIZER_VERSION" +#endif + +#if STBTT_RASTERIZER_VERSION == 1 +// note: this routine clips fills that extend off the edges... ideally this +// wouldn't happen, but it could happen if the truetype glyph bounding boxes +// are wrong, or if the user supplies a too-small bitmap +static void stbtt__fill_active_edges(unsigned char *scanline, int len, stbtt__active_edge *e, int max_weight) +{ + // non-zero winding fill + int x0=0, w=0; + + while (e) { + if (w == 0) { + // if we're currently at zero, we need to record the edge start point + x0 = e->x; w += e->direction; + } else { + int x1 = e->x; w += e->direction; + // if we went to zero, we need to draw + if (w == 0) { + int i = x0 >> STBTT_FIXSHIFT; + int j = x1 >> STBTT_FIXSHIFT; + + if (i < len && j >= 0) { + if (i == j) { + // x0,x1 are the same pixel, so compute combined coverage + scanline[i] = scanline[i] + (stbtt_uint8) ((x1 - x0) * max_weight >> STBTT_FIXSHIFT); + } else { + if (i >= 0) // add antialiasing for x0 + scanline[i] = scanline[i] + (stbtt_uint8) (((STBTT_FIX - (x0 & STBTT_FIXMASK)) * max_weight) >> STBTT_FIXSHIFT); + else + i = -1; // clip + + if (j < len) // add antialiasing for x1 + scanline[j] = scanline[j] + (stbtt_uint8) (((x1 & STBTT_FIXMASK) * max_weight) >> STBTT_FIXSHIFT); + else + j = len; // clip + + for (++i; i < j; ++i) // fill pixels between x0 and x1 + scanline[i] = scanline[i] + (stbtt_uint8) max_weight; + } + } + } + } + + e = e->next; + } +} + +static void stbtt__rasterize_sorted_edges(stbtt__bitmap *result, stbtt__edge *e, int n, int vsubsample, int off_x, int off_y, void *userdata) +{ + stbtt__hheap hh = { 0, 0, 0 }; + stbtt__active_edge *active = NULL; + int y,j=0; + int max_weight = (255 / vsubsample); // weight per vertical scanline + int s; // vertical subsample index + unsigned char scanline_data[512], *scanline; + + if (result->w > 512) + scanline = (unsigned char *) STBTT_malloc(result->w, userdata); + else + scanline = scanline_data; + + y = off_y * vsubsample; + e[n].y0 = (off_y + result->h) * (float) vsubsample + 1; + + while (j < result->h) { + STBTT_memset(scanline, 0, result->w); + for (s=0; s < vsubsample; ++s) { + // find center of pixel for this scanline + float scan_y = y + 0.5f; + stbtt__active_edge **step = &active; + + // update all active edges; + // remove all active edges that terminate before the center of this scanline + while (*step) { + stbtt__active_edge * z = *step; + if (z->ey <= scan_y) { + *step = z->next; // delete from list + STBTT_assert(z->direction); + z->direction = 0; + stbtt__hheap_free(&hh, z); + } else { + z->x += z->dx; // advance to position for current scanline + step = &((*step)->next); // advance through list + } + } + + // resort the list if needed + for(;;) { + int changed=0; + step = &active; + while (*step && (*step)->next) { + if ((*step)->x > (*step)->next->x) { + stbtt__active_edge *t = *step; + stbtt__active_edge *q = t->next; + + t->next = q->next; + q->next = t; + *step = q; + changed = 1; + } + step = &(*step)->next; + } + if (!changed) break; + } + + // insert all edges that start before the center of this scanline -- omit ones that also end on this scanline + while (e->y0 <= scan_y) { + if (e->y1 > scan_y) { + stbtt__active_edge *z = stbtt__new_active(&hh, e, off_x, scan_y, userdata); + if (z != NULL) { + // find insertion point + if (active == NULL) + active = z; + else if (z->x < active->x) { + // insert at front + z->next = active; + active = z; + } else { + // find thing to insert AFTER + stbtt__active_edge *p = active; + while (p->next && p->next->x < z->x) + p = p->next; + // at this point, p->next->x is NOT < z->x + z->next = p->next; + p->next = z; + } + } + } + ++e; + } + + // now process all active edges in XOR fashion + if (active) + stbtt__fill_active_edges(scanline, result->w, active, max_weight); + + ++y; + } + STBTT_memcpy(result->pixels + j * result->stride, scanline, result->w); + ++j; + } + + stbtt__hheap_cleanup(&hh, userdata); + + if (scanline != scanline_data) + STBTT_free(scanline, userdata); +} + +#elif STBTT_RASTERIZER_VERSION == 2 + +// the edge passed in here does not cross the vertical line at x or the vertical line at x+1 +// (i.e. it has already been clipped to those) +static void stbtt__handle_clipped_edge(float *scanline, int x, stbtt__active_edge *e, float x0, float y0, float x1, float y1) +{ + if (y0 == y1) return; + STBTT_assert(y0 < y1); + STBTT_assert(e->sy <= e->ey); + if (y0 > e->ey) return; + if (y1 < e->sy) return; + if (y0 < e->sy) { + x0 += (x1-x0) * (e->sy - y0) / (y1-y0); + y0 = e->sy; + } + if (y1 > e->ey) { + x1 += (x1-x0) * (e->ey - y1) / (y1-y0); + y1 = e->ey; + } + + if (x0 == x) + STBTT_assert(x1 <= x+1); + else if (x0 == x+1) + STBTT_assert(x1 >= x); + else if (x0 <= x) + STBTT_assert(x1 <= x); + else if (x0 >= x+1) + STBTT_assert(x1 >= x+1); + else + STBTT_assert(x1 >= x && x1 <= x+1); + + if (x0 <= x && x1 <= x) + scanline[x] += e->direction * (y1-y0); + else if (x0 >= x+1 && x1 >= x+1) + ; + else { + STBTT_assert(x0 >= x && x0 <= x+1 && x1 >= x && x1 <= x+1); + scanline[x] += e->direction * (y1-y0) * (1-((x0-x)+(x1-x))/2); // coverage = 1 - average x position + } +} + +static void stbtt__fill_active_edges_new(float *scanline, float *scanline_fill, int len, stbtt__active_edge *e, float y_top) +{ + float y_bottom = y_top+1; + + while (e) { + // brute force every pixel + + // compute intersection points with top & bottom + STBTT_assert(e->ey >= y_top); + + if (e->fdx == 0) { + float x0 = e->fx; + if (x0 < len) { + if (x0 >= 0) { + stbtt__handle_clipped_edge(scanline,(int) x0,e, x0,y_top, x0,y_bottom); + stbtt__handle_clipped_edge(scanline_fill-1,(int) x0+1,e, x0,y_top, x0,y_bottom); + } else { + stbtt__handle_clipped_edge(scanline_fill-1,0,e, x0,y_top, x0,y_bottom); + } + } + } else { + float x0 = e->fx; + float dx = e->fdx; + float xb = x0 + dx; + float x_top, x_bottom; + float sy0,sy1; + float dy = e->fdy; + STBTT_assert(e->sy <= y_bottom && e->ey >= y_top); + + // compute endpoints of line segment clipped to this scanline (if the + // line segment starts on this scanline. x0 is the intersection of the + // line with y_top, but that may be off the line segment. + if (e->sy > y_top) { + x_top = x0 + dx * (e->sy - y_top); + sy0 = e->sy; + } else { + x_top = x0; + sy0 = y_top; + } + if (e->ey < y_bottom) { + x_bottom = x0 + dx * (e->ey - y_top); + sy1 = e->ey; + } else { + x_bottom = xb; + sy1 = y_bottom; + } + + if (x_top >= 0 && x_bottom >= 0 && x_top < len && x_bottom < len) { + // from here on, we don't have to range check x values + + if ((int) x_top == (int) x_bottom) { + float height; + // simple case, only spans one pixel + int x = (int) x_top; + height = sy1 - sy0; + STBTT_assert(x >= 0 && x < len); + scanline[x] += e->direction * (1-((x_top - x) + (x_bottom-x))/2) * height; + scanline_fill[x] += e->direction * height; // everything right of this pixel is filled + } else { + int x,x1,x2; + float y_crossing, step, sign, area; + // covers 2+ pixels + if (x_top > x_bottom) { + // flip scanline vertically; signed area is the same + float t; + sy0 = y_bottom - (sy0 - y_top); + sy1 = y_bottom - (sy1 - y_top); + t = sy0, sy0 = sy1, sy1 = t; + t = x_bottom, x_bottom = x_top, x_top = t; + dx = -dx; + dy = -dy; + t = x0, x0 = xb, xb = t; + } + + x1 = (int) x_top; + x2 = (int) x_bottom; + // compute intersection with y axis at x1+1 + y_crossing = (x1+1 - x0) * dy + y_top; + + sign = e->direction; + // area of the rectangle covered from y0..y_crossing + area = sign * (y_crossing-sy0); + // area of the triangle (x_top,y0), (x+1,y0), (x+1,y_crossing) + scanline[x1] += area * (1-((x_top - x1)+(x1+1-x1))/2); + + step = sign * dy; + for (x = x1+1; x < x2; ++x) { + scanline[x] += area + step/2; + area += step; + } + y_crossing += dy * (x2 - (x1+1)); + + STBTT_assert(STBTT_fabs(area) <= 1.01f); + + scanline[x2] += area + sign * (1-((x2-x2)+(x_bottom-x2))/2) * (sy1-y_crossing); + + scanline_fill[x2] += sign * (sy1-sy0); + } + } else { + // if edge goes outside of box we're drawing, we require + // clipping logic. since this does not match the intended use + // of this library, we use a different, very slow brute + // force implementation + int x; + for (x=0; x < len; ++x) { + // cases: + // + // there can be up to two intersections with the pixel. any intersection + // with left or right edges can be handled by splitting into two (or three) + // regions. intersections with top & bottom do not necessitate case-wise logic. + // + // the old way of doing this found the intersections with the left & right edges, + // then used some simple logic to produce up to three segments in sorted order + // from top-to-bottom. however, this had a problem: if an x edge was epsilon + // across the x border, then the corresponding y position might not be distinct + // from the other y segment, and it might ignored as an empty segment. to avoid + // that, we need to explicitly produce segments based on x positions. + + // rename variables to clearly-defined pairs + float y0 = y_top; + float x1 = (float) (x); + float x2 = (float) (x+1); + float x3 = xb; + float y3 = y_bottom; + + // x = e->x + e->dx * (y-y_top) + // (y-y_top) = (x - e->x) / e->dx + // y = (x - e->x) / e->dx + y_top + float y1 = (x - x0) / dx + y_top; + float y2 = (x+1 - x0) / dx + y_top; + + if (x0 < x1 && x3 > x2) { // three segments descending down-right + stbtt__handle_clipped_edge(scanline,x,e, x0,y0, x1,y1); + stbtt__handle_clipped_edge(scanline,x,e, x1,y1, x2,y2); + stbtt__handle_clipped_edge(scanline,x,e, x2,y2, x3,y3); + } else if (x3 < x1 && x0 > x2) { // three segments descending down-left + stbtt__handle_clipped_edge(scanline,x,e, x0,y0, x2,y2); + stbtt__handle_clipped_edge(scanline,x,e, x2,y2, x1,y1); + stbtt__handle_clipped_edge(scanline,x,e, x1,y1, x3,y3); + } else if (x0 < x1 && x3 > x1) { // two segments across x, down-right + stbtt__handle_clipped_edge(scanline,x,e, x0,y0, x1,y1); + stbtt__handle_clipped_edge(scanline,x,e, x1,y1, x3,y3); + } else if (x3 < x1 && x0 > x1) { // two segments across x, down-left + stbtt__handle_clipped_edge(scanline,x,e, x0,y0, x1,y1); + stbtt__handle_clipped_edge(scanline,x,e, x1,y1, x3,y3); + } else if (x0 < x2 && x3 > x2) { // two segments across x+1, down-right + stbtt__handle_clipped_edge(scanline,x,e, x0,y0, x2,y2); + stbtt__handle_clipped_edge(scanline,x,e, x2,y2, x3,y3); + } else if (x3 < x2 && x0 > x2) { // two segments across x+1, down-left + stbtt__handle_clipped_edge(scanline,x,e, x0,y0, x2,y2); + stbtt__handle_clipped_edge(scanline,x,e, x2,y2, x3,y3); + } else { // one segment + stbtt__handle_clipped_edge(scanline,x,e, x0,y0, x3,y3); + } + } + } + } + e = e->next; + } +} + +// directly AA rasterize edges w/o supersampling +static void stbtt__rasterize_sorted_edges(stbtt__bitmap *result, stbtt__edge *e, int n, int vsubsample, int off_x, int off_y, void *userdata) +{ + stbtt__hheap hh = { 0, 0, 0 }; + stbtt__active_edge *active = NULL; + int y,j=0, i; + float scanline_data[129], *scanline, *scanline2; + + STBTT__NOTUSED(vsubsample); + + if (result->w > 64) + scanline = (float *) STBTT_malloc((result->w*2+1) * sizeof(float), userdata); + else + scanline = scanline_data; + + scanline2 = scanline + result->w; + + y = off_y; + e[n].y0 = (float) (off_y + result->h) + 1; + + while (j < result->h) { + // find center of pixel for this scanline + float scan_y_top = y + 0.0f; + float scan_y_bottom = y + 1.0f; + stbtt__active_edge **step = &active; + + STBTT_memset(scanline , 0, result->w*sizeof(scanline[0])); + STBTT_memset(scanline2, 0, (result->w+1)*sizeof(scanline[0])); + + // update all active edges; + // remove all active edges that terminate before the top of this scanline + while (*step) { + stbtt__active_edge * z = *step; + if (z->ey <= scan_y_top) { + *step = z->next; // delete from list + STBTT_assert(z->direction); + z->direction = 0; + stbtt__hheap_free(&hh, z); + } else { + step = &((*step)->next); // advance through list + } + } + + // insert all edges that start before the bottom of this scanline + while (e->y0 <= scan_y_bottom) { + if (e->y0 != e->y1) { + stbtt__active_edge *z = stbtt__new_active(&hh, e, off_x, scan_y_top, userdata); + if (z != NULL) { + if (j == 0 && off_y != 0) { + if (z->ey < scan_y_top) { + // this can happen due to subpixel positioning and some kind of fp rounding error i think + z->ey = scan_y_top; + } + } + STBTT_assert(z->ey >= scan_y_top); // if we get really unlucky a tiny bit of an edge can be out of bounds + // insert at front + z->next = active; + active = z; + } + } + ++e; + } + + // now process all active edges + if (active) + stbtt__fill_active_edges_new(scanline, scanline2+1, result->w, active, scan_y_top); + + { + float sum = 0; + for (i=0; i < result->w; ++i) { + float k; + int m; + sum += scanline2[i]; + k = scanline[i] + sum; + k = (float) STBTT_fabs(k)*255 + 0.5f; + m = (int) k; + if (m > 255) m = 255; + result->pixels[j*result->stride + i] = (unsigned char) m; + } + } + // advance all the edges + step = &active; + while (*step) { + stbtt__active_edge *z = *step; + z->fx += z->fdx; // advance to position for current scanline + step = &((*step)->next); // advance through list + } + + ++y; + ++j; + } + + stbtt__hheap_cleanup(&hh, userdata); + + if (scanline != scanline_data) + STBTT_free(scanline, userdata); +} +#else +#error "Unrecognized value of STBTT_RASTERIZER_VERSION" +#endif + +#define STBTT__COMPARE(a,b) ((a)->y0 < (b)->y0) + +static void stbtt__sort_edges_ins_sort(stbtt__edge *p, int n) +{ + int i,j; + for (i=1; i < n; ++i) { + stbtt__edge t = p[i], *a = &t; + j = i; + while (j > 0) { + stbtt__edge *b = &p[j-1]; + int c = STBTT__COMPARE(a,b); + if (!c) break; + p[j] = p[j-1]; + --j; + } + if (i != j) + p[j] = t; + } +} + +static void stbtt__sort_edges_quicksort(stbtt__edge *p, int n) +{ + /* threshold for transitioning to insertion sort */ + while (n > 12) { + stbtt__edge t; + int c01,c12,c,m,i,j; + + /* compute median of three */ + m = n >> 1; + c01 = STBTT__COMPARE(&p[0],&p[m]); + c12 = STBTT__COMPARE(&p[m],&p[n-1]); + /* if 0 >= mid >= end, or 0 < mid < end, then use mid */ + if (c01 != c12) { + /* otherwise, we'll need to swap something else to middle */ + int z; + c = STBTT__COMPARE(&p[0],&p[n-1]); + /* 0>mid && midn => n; 0 0 */ + /* 0n: 0>n => 0; 0 n */ + z = (c == c12) ? 0 : n-1; + t = p[z]; + p[z] = p[m]; + p[m] = t; + } + /* now p[m] is the median-of-three */ + /* swap it to the beginning so it won't move around */ + t = p[0]; + p[0] = p[m]; + p[m] = t; + + /* partition loop */ + i=1; + j=n-1; + for(;;) { + /* handling of equality is crucial here */ + /* for sentinels & efficiency with duplicates */ + for (;;++i) { + if (!STBTT__COMPARE(&p[i], &p[0])) break; + } + for (;;--j) { + if (!STBTT__COMPARE(&p[0], &p[j])) break; + } + /* make sure we haven't crossed */ + if (i >= j) break; + t = p[i]; + p[i] = p[j]; + p[j] = t; + + ++i; + --j; + } + /* recurse on smaller side, iterate on larger */ + if (j < (n-i)) { + stbtt__sort_edges_quicksort(p,j); + p = p+i; + n = n-i; + } else { + stbtt__sort_edges_quicksort(p+i, n-i); + n = j; + } + } +} + +static void stbtt__sort_edges(stbtt__edge *p, int n) +{ + stbtt__sort_edges_quicksort(p, n); + stbtt__sort_edges_ins_sort(p, n); +} + +typedef struct +{ + float x,y; +} stbtt__point; + +static void stbtt__rasterize(stbtt__bitmap *result, stbtt__point *pts, int *wcount, int windings, float scale_x, float scale_y, float shift_x, float shift_y, int off_x, int off_y, int invert, void *userdata) +{ + float y_scale_inv = invert ? -scale_y : scale_y; + stbtt__edge *e; + int n,i,j,k,m; +#if STBTT_RASTERIZER_VERSION == 1 + int vsubsample = result->h < 8 ? 15 : 5; +#elif STBTT_RASTERIZER_VERSION == 2 + int vsubsample = 1; +#else + #error "Unrecognized value of STBTT_RASTERIZER_VERSION" +#endif + // vsubsample should divide 255 evenly; otherwise we won't reach full opacity + + // now we have to blow out the windings into explicit edge lists + n = 0; + for (i=0; i < windings; ++i) + n += wcount[i]; + + e = (stbtt__edge *) STBTT_malloc(sizeof(*e) * (n+1), userdata); // add an extra one as a sentinel + if (e == 0) return; + n = 0; + + m=0; + for (i=0; i < windings; ++i) { + stbtt__point *p = pts + m; + m += wcount[i]; + j = wcount[i]-1; + for (k=0; k < wcount[i]; j=k++) { + int a=k,b=j; + // skip the edge if horizontal + if (p[j].y == p[k].y) + continue; + // add edge from j to k to the list + e[n].invert = 0; + if (invert ? p[j].y > p[k].y : p[j].y < p[k].y) { + e[n].invert = 1; + a=j,b=k; + } + e[n].x0 = p[a].x * scale_x + shift_x; + e[n].y0 = (p[a].y * y_scale_inv + shift_y) * vsubsample; + e[n].x1 = p[b].x * scale_x + shift_x; + e[n].y1 = (p[b].y * y_scale_inv + shift_y) * vsubsample; + ++n; + } + } + + // now sort the edges by their highest point (should snap to integer, and then by x) + //STBTT_sort(e, n, sizeof(e[0]), stbtt__edge_compare); + stbtt__sort_edges(e, n); + + // now, traverse the scanlines and find the intersections on each scanline, use xor winding rule + stbtt__rasterize_sorted_edges(result, e, n, vsubsample, off_x, off_y, userdata); + + STBTT_free(e, userdata); +} + +static void stbtt__add_point(stbtt__point *points, int n, float x, float y) +{ + if (!points) return; // during first pass, it's unallocated + points[n].x = x; + points[n].y = y; +} + +// tessellate until threshold p is happy... @TODO warped to compensate for non-linear stretching +static int stbtt__tesselate_curve(stbtt__point *points, int *num_points, float x0, float y0, float x1, float y1, float x2, float y2, float objspace_flatness_squared, int n) +{ + // midpoint + float mx = (x0 + 2*x1 + x2)/4; + float my = (y0 + 2*y1 + y2)/4; + // versus directly drawn line + float dx = (x0+x2)/2 - mx; + float dy = (y0+y2)/2 - my; + if (n > 16) // 65536 segments on one curve better be enough! + return 1; + if (dx*dx+dy*dy > objspace_flatness_squared) { // half-pixel error allowed... need to be smaller if AA + stbtt__tesselate_curve(points, num_points, x0,y0, (x0+x1)/2.0f,(y0+y1)/2.0f, mx,my, objspace_flatness_squared,n+1); + stbtt__tesselate_curve(points, num_points, mx,my, (x1+x2)/2.0f,(y1+y2)/2.0f, x2,y2, objspace_flatness_squared,n+1); + } else { + stbtt__add_point(points, *num_points,x2,y2); + *num_points = *num_points+1; + } + return 1; +} + +static void stbtt__tesselate_cubic(stbtt__point *points, int *num_points, float x0, float y0, float x1, float y1, float x2, float y2, float x3, float y3, float objspace_flatness_squared, int n) +{ + // @TODO this "flatness" calculation is just made-up nonsense that seems to work well enough + float dx0 = x1-x0; + float dy0 = y1-y0; + float dx1 = x2-x1; + float dy1 = y2-y1; + float dx2 = x3-x2; + float dy2 = y3-y2; + float dx = x3-x0; + float dy = y3-y0; + float longlen = (float) (STBTT_sqrt(dx0*dx0+dy0*dy0)+STBTT_sqrt(dx1*dx1+dy1*dy1)+STBTT_sqrt(dx2*dx2+dy2*dy2)); + float shortlen = (float) STBTT_sqrt(dx*dx+dy*dy); + float flatness_squared = longlen*longlen-shortlen*shortlen; + + if (n > 16) // 65536 segments on one curve better be enough! + return; + + if (flatness_squared > objspace_flatness_squared) { + float x01 = (x0+x1)/2; + float y01 = (y0+y1)/2; + float x12 = (x1+x2)/2; + float y12 = (y1+y2)/2; + float x23 = (x2+x3)/2; + float y23 = (y2+y3)/2; + + float xa = (x01+x12)/2; + float ya = (y01+y12)/2; + float xb = (x12+x23)/2; + float yb = (y12+y23)/2; + + float mx = (xa+xb)/2; + float my = (ya+yb)/2; + + stbtt__tesselate_cubic(points, num_points, x0,y0, x01,y01, xa,ya, mx,my, objspace_flatness_squared,n+1); + stbtt__tesselate_cubic(points, num_points, mx,my, xb,yb, x23,y23, x3,y3, objspace_flatness_squared,n+1); + } else { + stbtt__add_point(points, *num_points,x3,y3); + *num_points = *num_points+1; + } +} + +// returns number of contours +static stbtt__point *stbtt_FlattenCurves(stbtt_vertex *vertices, int num_verts, float objspace_flatness, int **contour_lengths, int *num_contours, void *userdata) +{ + stbtt__point *points=0; + int num_points=0; + + float objspace_flatness_squared = objspace_flatness * objspace_flatness; + int i,n=0,start=0, pass; + + // count how many "moves" there are to get the contour count + for (i=0; i < num_verts; ++i) + if (vertices[i].type == STBTT_vmove) + ++n; + + *num_contours = n; + if (n == 0) return 0; + + *contour_lengths = (int *) STBTT_malloc(sizeof(**contour_lengths) * n, userdata); + + if (*contour_lengths == 0) { + *num_contours = 0; + return 0; + } + + // make two passes through the points so we don't need to realloc + for (pass=0; pass < 2; ++pass) { + float x=0,y=0; + if (pass == 1) { + points = (stbtt__point *) STBTT_malloc(num_points * sizeof(points[0]), userdata); + if (points == NULL) goto error; + } + num_points = 0; + n= -1; + for (i=0; i < num_verts; ++i) { + switch (vertices[i].type) { + case STBTT_vmove: + // start the next contour + if (n >= 0) + (*contour_lengths)[n] = num_points - start; + ++n; + start = num_points; + + x = vertices[i].x, y = vertices[i].y; + stbtt__add_point(points, num_points++, x,y); + break; + case STBTT_vline: + x = vertices[i].x, y = vertices[i].y; + stbtt__add_point(points, num_points++, x, y); + break; + case STBTT_vcurve: + stbtt__tesselate_curve(points, &num_points, x,y, + vertices[i].cx, vertices[i].cy, + vertices[i].x, vertices[i].y, + objspace_flatness_squared, 0); + x = vertices[i].x, y = vertices[i].y; + break; + case STBTT_vcubic: + stbtt__tesselate_cubic(points, &num_points, x,y, + vertices[i].cx, vertices[i].cy, + vertices[i].cx1, vertices[i].cy1, + vertices[i].x, vertices[i].y, + objspace_flatness_squared, 0); + x = vertices[i].x, y = vertices[i].y; + break; + } + } + (*contour_lengths)[n] = num_points - start; + } + + return points; +error: + STBTT_free(points, userdata); + STBTT_free(*contour_lengths, userdata); + *contour_lengths = 0; + *num_contours = 0; + return NULL; +} + +STBTT_DEF void stbtt_Rasterize(stbtt__bitmap *result, float flatness_in_pixels, stbtt_vertex *vertices, int num_verts, float scale_x, float scale_y, float shift_x, float shift_y, int x_off, int y_off, int invert, void *userdata) +{ + float scale = scale_x > scale_y ? scale_y : scale_x; + int winding_count = 0; + int *winding_lengths = NULL; + stbtt__point *windings = stbtt_FlattenCurves(vertices, num_verts, flatness_in_pixels / scale, &winding_lengths, &winding_count, userdata); + if (windings) { + stbtt__rasterize(result, windings, winding_lengths, winding_count, scale_x, scale_y, shift_x, shift_y, x_off, y_off, invert, userdata); + STBTT_free(winding_lengths, userdata); + STBTT_free(windings, userdata); + } +} + +STBTT_DEF void stbtt_FreeBitmap(unsigned char *bitmap, void *userdata) +{ + STBTT_free(bitmap, userdata); +} + +STBTT_DEF unsigned char *stbtt_GetGlyphBitmapSubpixel(const stbtt_fontinfo *info, float scale_x, float scale_y, float shift_x, float shift_y, int glyph, int *width, int *height, int *xoff, int *yoff) +{ + int ix0,iy0,ix1,iy1; + stbtt__bitmap gbm; + stbtt_vertex *vertices; + int num_verts = stbtt_GetGlyphShape(info, glyph, &vertices); + + if (scale_x == 0) scale_x = scale_y; + if (scale_y == 0) { + if (scale_x == 0) { + STBTT_free(vertices, info->userdata); + return NULL; + } + scale_y = scale_x; + } + + stbtt_GetGlyphBitmapBoxSubpixel(info, glyph, scale_x, scale_y, shift_x, shift_y, &ix0,&iy0,&ix1,&iy1); + + // now we get the size + gbm.w = (ix1 - ix0); + gbm.h = (iy1 - iy0); + gbm.pixels = NULL; // in case we error + + if (width ) *width = gbm.w; + if (height) *height = gbm.h; + if (xoff ) *xoff = ix0; + if (yoff ) *yoff = iy0; + + if (gbm.w && gbm.h) { + gbm.pixels = (unsigned char *) STBTT_malloc(gbm.w * gbm.h, info->userdata); + if (gbm.pixels) { + gbm.stride = gbm.w; + + stbtt_Rasterize(&gbm, 0.35f, vertices, num_verts, scale_x, scale_y, shift_x, shift_y, ix0, iy0, 1, info->userdata); + } + } + STBTT_free(vertices, info->userdata); + return gbm.pixels; +} + +STBTT_DEF unsigned char *stbtt_GetGlyphBitmap(const stbtt_fontinfo *info, float scale_x, float scale_y, int glyph, int *width, int *height, int *xoff, int *yoff) +{ + return stbtt_GetGlyphBitmapSubpixel(info, scale_x, scale_y, 0.0f, 0.0f, glyph, width, height, xoff, yoff); +} + +STBTT_DEF void stbtt_MakeGlyphBitmapSubpixel(const stbtt_fontinfo *info, unsigned char *output, int out_w, int out_h, int out_stride, float scale_x, float scale_y, float shift_x, float shift_y, int glyph) +{ + int ix0,iy0; + stbtt_vertex *vertices; + int num_verts = stbtt_GetGlyphShape(info, glyph, &vertices); + stbtt__bitmap gbm; + + stbtt_GetGlyphBitmapBoxSubpixel(info, glyph, scale_x, scale_y, shift_x, shift_y, &ix0,&iy0,0,0); + gbm.pixels = output; + gbm.w = out_w; + gbm.h = out_h; + gbm.stride = out_stride; + + if (gbm.w && gbm.h) + stbtt_Rasterize(&gbm, 0.35f, vertices, num_verts, scale_x, scale_y, shift_x, shift_y, ix0,iy0, 1, info->userdata); + + STBTT_free(vertices, info->userdata); +} + +STBTT_DEF void stbtt_MakeGlyphBitmap(const stbtt_fontinfo *info, unsigned char *output, int out_w, int out_h, int out_stride, float scale_x, float scale_y, int glyph) +{ + stbtt_MakeGlyphBitmapSubpixel(info, output, out_w, out_h, out_stride, scale_x, scale_y, 0.0f,0.0f, glyph); +} + +STBTT_DEF unsigned char *stbtt_GetCodepointBitmapSubpixel(const stbtt_fontinfo *info, float scale_x, float scale_y, float shift_x, float shift_y, int codepoint, int *width, int *height, int *xoff, int *yoff) +{ + return stbtt_GetGlyphBitmapSubpixel(info, scale_x, scale_y,shift_x,shift_y, stbtt_FindGlyphIndex(info,codepoint), width,height,xoff,yoff); +} + +STBTT_DEF void stbtt_MakeCodepointBitmapSubpixelPrefilter(const stbtt_fontinfo *info, unsigned char *output, int out_w, int out_h, int out_stride, float scale_x, float scale_y, float shift_x, float shift_y, int oversample_x, int oversample_y, float *sub_x, float *sub_y, int codepoint) +{ + stbtt_MakeGlyphBitmapSubpixelPrefilter(info, output, out_w, out_h, out_stride, scale_x, scale_y, shift_x, shift_y, oversample_x, oversample_y, sub_x, sub_y, stbtt_FindGlyphIndex(info,codepoint)); +} + +STBTT_DEF void stbtt_MakeCodepointBitmapSubpixel(const stbtt_fontinfo *info, unsigned char *output, int out_w, int out_h, int out_stride, float scale_x, float scale_y, float shift_x, float shift_y, int codepoint) +{ + stbtt_MakeGlyphBitmapSubpixel(info, output, out_w, out_h, out_stride, scale_x, scale_y, shift_x, shift_y, stbtt_FindGlyphIndex(info,codepoint)); +} + +STBTT_DEF unsigned char *stbtt_GetCodepointBitmap(const stbtt_fontinfo *info, float scale_x, float scale_y, int codepoint, int *width, int *height, int *xoff, int *yoff) +{ + return stbtt_GetCodepointBitmapSubpixel(info, scale_x, scale_y, 0.0f,0.0f, codepoint, width,height,xoff,yoff); +} + +STBTT_DEF void stbtt_MakeCodepointBitmap(const stbtt_fontinfo *info, unsigned char *output, int out_w, int out_h, int out_stride, float scale_x, float scale_y, int codepoint) +{ + stbtt_MakeCodepointBitmapSubpixel(info, output, out_w, out_h, out_stride, scale_x, scale_y, 0.0f,0.0f, codepoint); +} + +////////////////////////////////////////////////////////////////////////////// +// +// bitmap baking +// +// This is SUPER-CRAPPY packing to keep source code small + +static int stbtt_BakeFontBitmap_internal(unsigned char *data, int offset, // font location (use offset=0 for plain .ttf) + float pixel_height, // height of font in pixels + unsigned char *pixels, int pw, int ph, // bitmap to be filled in + int first_char, int num_chars, // characters to bake + stbtt_bakedchar *chardata) +{ + float scale; + int x,y,bottom_y, i; + stbtt_fontinfo f; + f.userdata = NULL; + if (!stbtt_InitFont(&f, data, offset)) + return -1; + STBTT_memset(pixels, 0, pw*ph); // background of 0 around pixels + x=y=1; + bottom_y = 1; + + scale = stbtt_ScaleForPixelHeight(&f, pixel_height); + + for (i=0; i < num_chars; ++i) { + int advance, lsb, x0,y0,x1,y1,gw,gh; + int g = stbtt_FindGlyphIndex(&f, first_char + i); + stbtt_GetGlyphHMetrics(&f, g, &advance, &lsb); + stbtt_GetGlyphBitmapBox(&f, g, scale,scale, &x0,&y0,&x1,&y1); + gw = x1-x0; + gh = y1-y0; + if (x + gw + 1 >= pw) + y = bottom_y, x = 1; // advance to next row + if (y + gh + 1 >= ph) // check if it fits vertically AFTER potentially moving to next row + return -i; + STBTT_assert(x+gw < pw); + STBTT_assert(y+gh < ph); + stbtt_MakeGlyphBitmap(&f, pixels+x+y*pw, gw,gh,pw, scale,scale, g); + chardata[i].x0 = (stbtt_int16) x; + chardata[i].y0 = (stbtt_int16) y; + chardata[i].x1 = (stbtt_int16) (x + gw); + chardata[i].y1 = (stbtt_int16) (y + gh); + chardata[i].xadvance = scale * advance; + chardata[i].xoff = (float) x0; + chardata[i].yoff = (float) y0; + x = x + gw + 1; + if (y+gh+1 > bottom_y) + bottom_y = y+gh+1; + } + return bottom_y; +} + +STBTT_DEF void stbtt_GetBakedQuad(const stbtt_bakedchar *chardata, int pw, int ph, int char_index, float *xpos, float *ypos, stbtt_aligned_quad *q, int opengl_fillrule) +{ + float d3d_bias = opengl_fillrule ? 0 : -0.5f; + float ipw = 1.0f / pw, iph = 1.0f / ph; + const stbtt_bakedchar *b = chardata + char_index; + int round_x = STBTT_ifloor((*xpos + b->xoff) + 0.5f); + int round_y = STBTT_ifloor((*ypos + b->yoff) + 0.5f); + + q->x0 = round_x + d3d_bias; + q->y0 = round_y + d3d_bias; + q->x1 = round_x + b->x1 - b->x0 + d3d_bias; + q->y1 = round_y + b->y1 - b->y0 + d3d_bias; + + q->s0 = b->x0 * ipw; + q->t0 = b->y0 * iph; + q->s1 = b->x1 * ipw; + q->t1 = b->y1 * iph; + + *xpos += b->xadvance; +} + +////////////////////////////////////////////////////////////////////////////// +// +// rectangle packing replacement routines if you don't have stb_rect_pack.h +// + +#ifndef STB_RECT_PACK_VERSION + +typedef int stbrp_coord; + +//////////////////////////////////////////////////////////////////////////////////// +// // +// // +// COMPILER WARNING ?!?!? // +// // +// // +// if you get a compile warning due to these symbols being defined more than // +// once, move #include "stb_rect_pack.h" before #include "stb_truetype.h" // +// // +//////////////////////////////////////////////////////////////////////////////////// + +typedef struct +{ + int width,height; + int x,y,bottom_y; +} stbrp_context; + +typedef struct +{ + unsigned char x; +} stbrp_node; + +struct stbrp_rect +{ + stbrp_coord x,y; + int id,w,h,was_packed; +}; + +static void stbrp_init_target(stbrp_context *con, int pw, int ph, stbrp_node *nodes, int num_nodes) +{ + con->width = pw; + con->height = ph; + con->x = 0; + con->y = 0; + con->bottom_y = 0; + STBTT__NOTUSED(nodes); + STBTT__NOTUSED(num_nodes); +} + +static void stbrp_pack_rects(stbrp_context *con, stbrp_rect *rects, int num_rects) +{ + int i; + for (i=0; i < num_rects; ++i) { + if (con->x + rects[i].w > con->width) { + con->x = 0; + con->y = con->bottom_y; + } + if (con->y + rects[i].h > con->height) + break; + rects[i].x = con->x; + rects[i].y = con->y; + rects[i].was_packed = 1; + con->x += rects[i].w; + if (con->y + rects[i].h > con->bottom_y) + con->bottom_y = con->y + rects[i].h; + } + for ( ; i < num_rects; ++i) + rects[i].was_packed = 0; +} +#endif + +////////////////////////////////////////////////////////////////////////////// +// +// bitmap baking +// +// This is SUPER-AWESOME (tm Ryan Gordon) packing using stb_rect_pack.h. If +// stb_rect_pack.h isn't available, it uses the BakeFontBitmap strategy. + +STBTT_DEF int stbtt_PackBegin(stbtt_pack_context *spc, unsigned char *pixels, int pw, int ph, int stride_in_bytes, int padding, void *alloc_context) +{ + stbrp_context *context = (stbrp_context *) STBTT_malloc(sizeof(*context) ,alloc_context); + int num_nodes = pw - padding; + stbrp_node *nodes = (stbrp_node *) STBTT_malloc(sizeof(*nodes ) * num_nodes,alloc_context); + + if (context == NULL || nodes == NULL) { + if (context != NULL) STBTT_free(context, alloc_context); + if (nodes != NULL) STBTT_free(nodes , alloc_context); + return 0; + } + + spc->user_allocator_context = alloc_context; + spc->width = pw; + spc->height = ph; + spc->pixels = pixels; + spc->pack_info = context; + spc->nodes = nodes; + spc->padding = padding; + spc->stride_in_bytes = stride_in_bytes != 0 ? stride_in_bytes : pw; + spc->h_oversample = 1; + spc->v_oversample = 1; + spc->skip_missing = 0; + + stbrp_init_target(context, pw-padding, ph-padding, nodes, num_nodes); + + if (pixels) + STBTT_memset(pixels, 0, pw*ph); // background of 0 around pixels + + return 1; +} + +STBTT_DEF void stbtt_PackEnd (stbtt_pack_context *spc) +{ + STBTT_free(spc->nodes , spc->user_allocator_context); + STBTT_free(spc->pack_info, spc->user_allocator_context); +} + +STBTT_DEF void stbtt_PackSetOversampling(stbtt_pack_context *spc, unsigned int h_oversample, unsigned int v_oversample) +{ + STBTT_assert(h_oversample <= STBTT_MAX_OVERSAMPLE); + STBTT_assert(v_oversample <= STBTT_MAX_OVERSAMPLE); + if (h_oversample <= STBTT_MAX_OVERSAMPLE) + spc->h_oversample = h_oversample; + if (v_oversample <= STBTT_MAX_OVERSAMPLE) + spc->v_oversample = v_oversample; +} + +STBTT_DEF void stbtt_PackSetSkipMissingCodepoints(stbtt_pack_context *spc, int skip) +{ + spc->skip_missing = skip; +} + +#define STBTT__OVER_MASK (STBTT_MAX_OVERSAMPLE-1) + +static void stbtt__h_prefilter(unsigned char *pixels, int w, int h, int stride_in_bytes, unsigned int kernel_width) +{ + unsigned char buffer[STBTT_MAX_OVERSAMPLE]; + int safe_w = w - kernel_width; + int j; + STBTT_memset(buffer, 0, STBTT_MAX_OVERSAMPLE); // suppress bogus warning from VS2013 -analyze + for (j=0; j < h; ++j) { + int i; + unsigned int total; + STBTT_memset(buffer, 0, kernel_width); + + total = 0; + + // make kernel_width a constant in common cases so compiler can optimize out the divide + switch (kernel_width) { + case 2: + for (i=0; i <= safe_w; ++i) { + total += pixels[i] - buffer[i & STBTT__OVER_MASK]; + buffer[(i+kernel_width) & STBTT__OVER_MASK] = pixels[i]; + pixels[i] = (unsigned char) (total / 2); + } + break; + case 3: + for (i=0; i <= safe_w; ++i) { + total += pixels[i] - buffer[i & STBTT__OVER_MASK]; + buffer[(i+kernel_width) & STBTT__OVER_MASK] = pixels[i]; + pixels[i] = (unsigned char) (total / 3); + } + break; + case 4: + for (i=0; i <= safe_w; ++i) { + total += pixels[i] - buffer[i & STBTT__OVER_MASK]; + buffer[(i+kernel_width) & STBTT__OVER_MASK] = pixels[i]; + pixels[i] = (unsigned char) (total / 4); + } + break; + case 5: + for (i=0; i <= safe_w; ++i) { + total += pixels[i] - buffer[i & STBTT__OVER_MASK]; + buffer[(i+kernel_width) & STBTT__OVER_MASK] = pixels[i]; + pixels[i] = (unsigned char) (total / 5); + } + break; + default: + for (i=0; i <= safe_w; ++i) { + total += pixels[i] - buffer[i & STBTT__OVER_MASK]; + buffer[(i+kernel_width) & STBTT__OVER_MASK] = pixels[i]; + pixels[i] = (unsigned char) (total / kernel_width); + } + break; + } + + for (; i < w; ++i) { + STBTT_assert(pixels[i] == 0); + total -= buffer[i & STBTT__OVER_MASK]; + pixels[i] = (unsigned char) (total / kernel_width); + } + + pixels += stride_in_bytes; + } +} + +static void stbtt__v_prefilter(unsigned char *pixels, int w, int h, int stride_in_bytes, unsigned int kernel_width) +{ + unsigned char buffer[STBTT_MAX_OVERSAMPLE]; + int safe_h = h - kernel_width; + int j; + STBTT_memset(buffer, 0, STBTT_MAX_OVERSAMPLE); // suppress bogus warning from VS2013 -analyze + for (j=0; j < w; ++j) { + int i; + unsigned int total; + STBTT_memset(buffer, 0, kernel_width); + + total = 0; + + // make kernel_width a constant in common cases so compiler can optimize out the divide + switch (kernel_width) { + case 2: + for (i=0; i <= safe_h; ++i) { + total += pixels[i*stride_in_bytes] - buffer[i & STBTT__OVER_MASK]; + buffer[(i+kernel_width) & STBTT__OVER_MASK] = pixels[i*stride_in_bytes]; + pixels[i*stride_in_bytes] = (unsigned char) (total / 2); + } + break; + case 3: + for (i=0; i <= safe_h; ++i) { + total += pixels[i*stride_in_bytes] - buffer[i & STBTT__OVER_MASK]; + buffer[(i+kernel_width) & STBTT__OVER_MASK] = pixels[i*stride_in_bytes]; + pixels[i*stride_in_bytes] = (unsigned char) (total / 3); + } + break; + case 4: + for (i=0; i <= safe_h; ++i) { + total += pixels[i*stride_in_bytes] - buffer[i & STBTT__OVER_MASK]; + buffer[(i+kernel_width) & STBTT__OVER_MASK] = pixels[i*stride_in_bytes]; + pixels[i*stride_in_bytes] = (unsigned char) (total / 4); + } + break; + case 5: + for (i=0; i <= safe_h; ++i) { + total += pixels[i*stride_in_bytes] - buffer[i & STBTT__OVER_MASK]; + buffer[(i+kernel_width) & STBTT__OVER_MASK] = pixels[i*stride_in_bytes]; + pixels[i*stride_in_bytes] = (unsigned char) (total / 5); + } + break; + default: + for (i=0; i <= safe_h; ++i) { + total += pixels[i*stride_in_bytes] - buffer[i & STBTT__OVER_MASK]; + buffer[(i+kernel_width) & STBTT__OVER_MASK] = pixels[i*stride_in_bytes]; + pixels[i*stride_in_bytes] = (unsigned char) (total / kernel_width); + } + break; + } + + for (; i < h; ++i) { + STBTT_assert(pixels[i*stride_in_bytes] == 0); + total -= buffer[i & STBTT__OVER_MASK]; + pixels[i*stride_in_bytes] = (unsigned char) (total / kernel_width); + } + + pixels += 1; + } +} + +static float stbtt__oversample_shift(int oversample) +{ + if (!oversample) + return 0.0f; + + // The prefilter is a box filter of width "oversample", + // which shifts phase by (oversample - 1)/2 pixels in + // oversampled space. We want to shift in the opposite + // direction to counter this. + return (float)-(oversample - 1) / (2.0f * (float)oversample); +} + +// rects array must be big enough to accommodate all characters in the given ranges +STBTT_DEF int stbtt_PackFontRangesGatherRects(stbtt_pack_context *spc, const stbtt_fontinfo *info, stbtt_pack_range *ranges, int num_ranges, stbrp_rect *rects) +{ + int i,j,k; + int missing_glyph_added = 0; + + k=0; + for (i=0; i < num_ranges; ++i) { + float fh = ranges[i].font_size; + float scale = fh > 0 ? stbtt_ScaleForPixelHeight(info, fh) : stbtt_ScaleForMappingEmToPixels(info, -fh); + ranges[i].h_oversample = (unsigned char) spc->h_oversample; + ranges[i].v_oversample = (unsigned char) spc->v_oversample; + for (j=0; j < ranges[i].num_chars; ++j) { + int x0,y0,x1,y1; + int codepoint = ranges[i].array_of_unicode_codepoints == NULL ? ranges[i].first_unicode_codepoint_in_range + j : ranges[i].array_of_unicode_codepoints[j]; + int glyph = stbtt_FindGlyphIndex(info, codepoint); + if (glyph == 0 && (spc->skip_missing || missing_glyph_added)) { + rects[k].w = rects[k].h = 0; + } else { + stbtt_GetGlyphBitmapBoxSubpixel(info,glyph, + scale * spc->h_oversample, + scale * spc->v_oversample, + 0,0, + &x0,&y0,&x1,&y1); + rects[k].w = (stbrp_coord) (x1-x0 + spc->padding + spc->h_oversample-1); + rects[k].h = (stbrp_coord) (y1-y0 + spc->padding + spc->v_oversample-1); + if (glyph == 0) + missing_glyph_added = 1; + } + ++k; + } + } + + return k; +} + +STBTT_DEF void stbtt_MakeGlyphBitmapSubpixelPrefilter(const stbtt_fontinfo *info, unsigned char *output, int out_w, int out_h, int out_stride, float scale_x, float scale_y, float shift_x, float shift_y, int prefilter_x, int prefilter_y, float *sub_x, float *sub_y, int glyph) +{ + stbtt_MakeGlyphBitmapSubpixel(info, + output, + out_w - (prefilter_x - 1), + out_h - (prefilter_y - 1), + out_stride, + scale_x, + scale_y, + shift_x, + shift_y, + glyph); + + if (prefilter_x > 1) + stbtt__h_prefilter(output, out_w, out_h, out_stride, prefilter_x); + + if (prefilter_y > 1) + stbtt__v_prefilter(output, out_w, out_h, out_stride, prefilter_y); + + *sub_x = stbtt__oversample_shift(prefilter_x); + *sub_y = stbtt__oversample_shift(prefilter_y); +} + +// rects array must be big enough to accommodate all characters in the given ranges +STBTT_DEF int stbtt_PackFontRangesRenderIntoRects(stbtt_pack_context *spc, const stbtt_fontinfo *info, stbtt_pack_range *ranges, int num_ranges, stbrp_rect *rects) +{ + int i,j,k, missing_glyph = -1, return_value = 1; + + // save current values + int old_h_over = spc->h_oversample; + int old_v_over = spc->v_oversample; + + k = 0; + for (i=0; i < num_ranges; ++i) { + float fh = ranges[i].font_size; + float scale = fh > 0 ? stbtt_ScaleForPixelHeight(info, fh) : stbtt_ScaleForMappingEmToPixels(info, -fh); + float recip_h,recip_v,sub_x,sub_y; + spc->h_oversample = ranges[i].h_oversample; + spc->v_oversample = ranges[i].v_oversample; + recip_h = 1.0f / spc->h_oversample; + recip_v = 1.0f / spc->v_oversample; + sub_x = stbtt__oversample_shift(spc->h_oversample); + sub_y = stbtt__oversample_shift(spc->v_oversample); + for (j=0; j < ranges[i].num_chars; ++j) { + stbrp_rect *r = &rects[k]; + if (r->was_packed && r->w != 0 && r->h != 0) { + stbtt_packedchar *bc = &ranges[i].chardata_for_range[j]; + int advance, lsb, x0,y0,x1,y1; + int codepoint = ranges[i].array_of_unicode_codepoints == NULL ? ranges[i].first_unicode_codepoint_in_range + j : ranges[i].array_of_unicode_codepoints[j]; + int glyph = stbtt_FindGlyphIndex(info, codepoint); + stbrp_coord pad = (stbrp_coord) spc->padding; + + // pad on left and top + r->x += pad; + r->y += pad; + r->w -= pad; + r->h -= pad; + stbtt_GetGlyphHMetrics(info, glyph, &advance, &lsb); + stbtt_GetGlyphBitmapBox(info, glyph, + scale * spc->h_oversample, + scale * spc->v_oversample, + &x0,&y0,&x1,&y1); + stbtt_MakeGlyphBitmapSubpixel(info, + spc->pixels + r->x + r->y*spc->stride_in_bytes, + r->w - spc->h_oversample+1, + r->h - spc->v_oversample+1, + spc->stride_in_bytes, + scale * spc->h_oversample, + scale * spc->v_oversample, + 0,0, + glyph); + + if (spc->h_oversample > 1) + stbtt__h_prefilter(spc->pixels + r->x + r->y*spc->stride_in_bytes, + r->w, r->h, spc->stride_in_bytes, + spc->h_oversample); + + if (spc->v_oversample > 1) + stbtt__v_prefilter(spc->pixels + r->x + r->y*spc->stride_in_bytes, + r->w, r->h, spc->stride_in_bytes, + spc->v_oversample); + + bc->x0 = (stbtt_int16) r->x; + bc->y0 = (stbtt_int16) r->y; + bc->x1 = (stbtt_int16) (r->x + r->w); + bc->y1 = (stbtt_int16) (r->y + r->h); + bc->xadvance = scale * advance; + bc->xoff = (float) x0 * recip_h + sub_x; + bc->yoff = (float) y0 * recip_v + sub_y; + bc->xoff2 = (x0 + r->w) * recip_h + sub_x; + bc->yoff2 = (y0 + r->h) * recip_v + sub_y; + + if (glyph == 0) + missing_glyph = j; + } else if (spc->skip_missing) { + return_value = 0; + } else if (r->was_packed && r->w == 0 && r->h == 0 && missing_glyph >= 0) { + ranges[i].chardata_for_range[j] = ranges[i].chardata_for_range[missing_glyph]; + } else { + return_value = 0; // if any fail, report failure + } + + ++k; + } + } + + // restore original values + spc->h_oversample = old_h_over; + spc->v_oversample = old_v_over; + + return return_value; +} + +STBTT_DEF void stbtt_PackFontRangesPackRects(stbtt_pack_context *spc, stbrp_rect *rects, int num_rects) +{ + stbrp_pack_rects((stbrp_context *) spc->pack_info, rects, num_rects); +} + +STBTT_DEF int stbtt_PackFontRanges(stbtt_pack_context *spc, const unsigned char *fontdata, int font_index, stbtt_pack_range *ranges, int num_ranges) +{ + stbtt_fontinfo info; + int i,j,n, return_value = 1; + //stbrp_context *context = (stbrp_context *) spc->pack_info; + stbrp_rect *rects; + + // flag all characters as NOT packed + for (i=0; i < num_ranges; ++i) + for (j=0; j < ranges[i].num_chars; ++j) + ranges[i].chardata_for_range[j].x0 = + ranges[i].chardata_for_range[j].y0 = + ranges[i].chardata_for_range[j].x1 = + ranges[i].chardata_for_range[j].y1 = 0; + + n = 0; + for (i=0; i < num_ranges; ++i) + n += ranges[i].num_chars; + + rects = (stbrp_rect *) STBTT_malloc(sizeof(*rects) * n, spc->user_allocator_context); + if (rects == NULL) + return 0; + + info.userdata = spc->user_allocator_context; + stbtt_InitFont(&info, fontdata, stbtt_GetFontOffsetForIndex(fontdata,font_index)); + + n = stbtt_PackFontRangesGatherRects(spc, &info, ranges, num_ranges, rects); + + stbtt_PackFontRangesPackRects(spc, rects, n); + + return_value = stbtt_PackFontRangesRenderIntoRects(spc, &info, ranges, num_ranges, rects); + + STBTT_free(rects, spc->user_allocator_context); + return return_value; +} + +STBTT_DEF int stbtt_PackFontRange(stbtt_pack_context *spc, const unsigned char *fontdata, int font_index, float font_size, + int first_unicode_codepoint_in_range, int num_chars_in_range, stbtt_packedchar *chardata_for_range) +{ + stbtt_pack_range range; + range.first_unicode_codepoint_in_range = first_unicode_codepoint_in_range; + range.array_of_unicode_codepoints = NULL; + range.num_chars = num_chars_in_range; + range.chardata_for_range = chardata_for_range; + range.font_size = font_size; + return stbtt_PackFontRanges(spc, fontdata, font_index, &range, 1); +} + +STBTT_DEF void stbtt_GetScaledFontVMetrics(const unsigned char *fontdata, int index, float size, float *ascent, float *descent, float *lineGap) +{ + int i_ascent, i_descent, i_lineGap; + float scale; + stbtt_fontinfo info; + stbtt_InitFont(&info, fontdata, stbtt_GetFontOffsetForIndex(fontdata, index)); + scale = size > 0 ? stbtt_ScaleForPixelHeight(&info, size) : stbtt_ScaleForMappingEmToPixels(&info, -size); + stbtt_GetFontVMetrics(&info, &i_ascent, &i_descent, &i_lineGap); + *ascent = (float) i_ascent * scale; + *descent = (float) i_descent * scale; + *lineGap = (float) i_lineGap * scale; +} + +STBTT_DEF void stbtt_GetPackedQuad(const stbtt_packedchar *chardata, int pw, int ph, int char_index, float *xpos, float *ypos, stbtt_aligned_quad *q, int align_to_integer) +{ + float ipw = 1.0f / pw, iph = 1.0f / ph; + const stbtt_packedchar *b = chardata + char_index; + + if (align_to_integer) { + float x = (float) STBTT_ifloor((*xpos + b->xoff) + 0.5f); + float y = (float) STBTT_ifloor((*ypos + b->yoff) + 0.5f); + q->x0 = x; + q->y0 = y; + q->x1 = x + b->xoff2 - b->xoff; + q->y1 = y + b->yoff2 - b->yoff; + } else { + q->x0 = *xpos + b->xoff; + q->y0 = *ypos + b->yoff; + q->x1 = *xpos + b->xoff2; + q->y1 = *ypos + b->yoff2; + } + + q->s0 = b->x0 * ipw; + q->t0 = b->y0 * iph; + q->s1 = b->x1 * ipw; + q->t1 = b->y1 * iph; + + *xpos += b->xadvance; +} + +////////////////////////////////////////////////////////////////////////////// +// +// sdf computation +// + +#define STBTT_min(a,b) ((a) < (b) ? (a) : (b)) +#define STBTT_max(a,b) ((a) < (b) ? (b) : (a)) + +static int stbtt__ray_intersect_bezier(float orig[2], float ray[2], float q0[2], float q1[2], float q2[2], float hits[2][2]) +{ + float q0perp = q0[1]*ray[0] - q0[0]*ray[1]; + float q1perp = q1[1]*ray[0] - q1[0]*ray[1]; + float q2perp = q2[1]*ray[0] - q2[0]*ray[1]; + float roperp = orig[1]*ray[0] - orig[0]*ray[1]; + + float a = q0perp - 2*q1perp + q2perp; + float b = q1perp - q0perp; + float c = q0perp - roperp; + + float s0 = 0., s1 = 0.; + int num_s = 0; + + if (a != 0.0) { + float discr = b*b - a*c; + if (discr > 0.0) { + float rcpna = -1 / a; + float d = (float) STBTT_sqrt(discr); + s0 = (b+d) * rcpna; + s1 = (b-d) * rcpna; + if (s0 >= 0.0 && s0 <= 1.0) + num_s = 1; + if (d > 0.0 && s1 >= 0.0 && s1 <= 1.0) { + if (num_s == 0) s0 = s1; + ++num_s; + } + } + } else { + // 2*b*s + c = 0 + // s = -c / (2*b) + s0 = c / (-2 * b); + if (s0 >= 0.0 && s0 <= 1.0) + num_s = 1; + } + + if (num_s == 0) + return 0; + else { + float rcp_len2 = 1 / (ray[0]*ray[0] + ray[1]*ray[1]); + float rayn_x = ray[0] * rcp_len2, rayn_y = ray[1] * rcp_len2; + + float q0d = q0[0]*rayn_x + q0[1]*rayn_y; + float q1d = q1[0]*rayn_x + q1[1]*rayn_y; + float q2d = q2[0]*rayn_x + q2[1]*rayn_y; + float rod = orig[0]*rayn_x + orig[1]*rayn_y; + + float q10d = q1d - q0d; + float q20d = q2d - q0d; + float q0rd = q0d - rod; + + hits[0][0] = q0rd + s0*(2.0f - 2.0f*s0)*q10d + s0*s0*q20d; + hits[0][1] = a*s0+b; + + if (num_s > 1) { + hits[1][0] = q0rd + s1*(2.0f - 2.0f*s1)*q10d + s1*s1*q20d; + hits[1][1] = a*s1+b; + return 2; + } else { + return 1; + } + } +} + +static int equal(float *a, float *b) +{ + return (a[0] == b[0] && a[1] == b[1]); +} + +static int stbtt__compute_crossings_x(float x, float y, int nverts, stbtt_vertex *verts) +{ + int i; + float orig[2], ray[2] = { 1, 0 }; + float y_frac; + int winding = 0; + + orig[0] = x; + orig[1] = y; + + // make sure y never passes through a vertex of the shape + y_frac = (float) STBTT_fmod(y, 1.0f); + if (y_frac < 0.01f) + y += 0.01f; + else if (y_frac > 0.99f) + y -= 0.01f; + orig[1] = y; + + // test a ray from (-infinity,y) to (x,y) + for (i=0; i < nverts; ++i) { + if (verts[i].type == STBTT_vline) { + int x0 = (int) verts[i-1].x, y0 = (int) verts[i-1].y; + int x1 = (int) verts[i ].x, y1 = (int) verts[i ].y; + if (y > STBTT_min(y0,y1) && y < STBTT_max(y0,y1) && x > STBTT_min(x0,x1)) { + float x_inter = (y - y0) / (y1 - y0) * (x1-x0) + x0; + if (x_inter < x) + winding += (y0 < y1) ? 1 : -1; + } + } + if (verts[i].type == STBTT_vcurve) { + int x0 = (int) verts[i-1].x , y0 = (int) verts[i-1].y ; + int x1 = (int) verts[i ].cx, y1 = (int) verts[i ].cy; + int x2 = (int) verts[i ].x , y2 = (int) verts[i ].y ; + int ax = STBTT_min(x0,STBTT_min(x1,x2)), ay = STBTT_min(y0,STBTT_min(y1,y2)); + int by = STBTT_max(y0,STBTT_max(y1,y2)); + if (y > ay && y < by && x > ax) { + float q0[2],q1[2],q2[2]; + float hits[2][2]; + q0[0] = (float)x0; + q0[1] = (float)y0; + q1[0] = (float)x1; + q1[1] = (float)y1; + q2[0] = (float)x2; + q2[1] = (float)y2; + if (equal(q0,q1) || equal(q1,q2)) { + x0 = (int)verts[i-1].x; + y0 = (int)verts[i-1].y; + x1 = (int)verts[i ].x; + y1 = (int)verts[i ].y; + if (y > STBTT_min(y0,y1) && y < STBTT_max(y0,y1) && x > STBTT_min(x0,x1)) { + float x_inter = (y - y0) / (y1 - y0) * (x1-x0) + x0; + if (x_inter < x) + winding += (y0 < y1) ? 1 : -1; + } + } else { + int num_hits = stbtt__ray_intersect_bezier(orig, ray, q0, q1, q2, hits); + if (num_hits >= 1) + if (hits[0][0] < 0) + winding += (hits[0][1] < 0 ? -1 : 1); + if (num_hits >= 2) + if (hits[1][0] < 0) + winding += (hits[1][1] < 0 ? -1 : 1); + } + } + } + } + return winding; +} + +static float stbtt__cuberoot( float x ) +{ + if (x<0) + return -(float) STBTT_pow(-x,1.0f/3.0f); + else + return (float) STBTT_pow( x,1.0f/3.0f); +} + +// x^3 + c*x^2 + b*x + a = 0 +static int stbtt__solve_cubic(float a, float b, float c, float* r) +{ + float s = -a / 3; + float p = b - a*a / 3; + float q = a * (2*a*a - 9*b) / 27 + c; + float p3 = p*p*p; + float d = q*q + 4*p3 / 27; + if (d >= 0) { + float z = (float) STBTT_sqrt(d); + float u = (-q + z) / 2; + float v = (-q - z) / 2; + u = stbtt__cuberoot(u); + v = stbtt__cuberoot(v); + r[0] = s + u + v; + return 1; + } else { + float u = (float) STBTT_sqrt(-p/3); + float v = (float) STBTT_acos(-STBTT_sqrt(-27/p3) * q / 2) / 3; // p3 must be negative, since d is negative + float m = (float) STBTT_cos(v); + float n = (float) STBTT_cos(v-3.141592/2)*1.732050808f; + r[0] = s + u * 2 * m; + r[1] = s - u * (m + n); + r[2] = s - u * (m - n); + + //STBTT_assert( STBTT_fabs(((r[0]+a)*r[0]+b)*r[0]+c) < 0.05f); // these asserts may not be safe at all scales, though they're in bezier t parameter units so maybe? + //STBTT_assert( STBTT_fabs(((r[1]+a)*r[1]+b)*r[1]+c) < 0.05f); + //STBTT_assert( STBTT_fabs(((r[2]+a)*r[2]+b)*r[2]+c) < 0.05f); + return 3; + } +} + +STBTT_DEF unsigned char * stbtt_GetGlyphSDF(const stbtt_fontinfo *info, float scale, int glyph, int padding, unsigned char onedge_value, float pixel_dist_scale, int *width, int *height, int *xoff, int *yoff) +{ + float scale_x = scale, scale_y = scale; + int ix0,iy0,ix1,iy1; + int w,h; + unsigned char *data; + + if (scale == 0) return NULL; + + stbtt_GetGlyphBitmapBoxSubpixel(info, glyph, scale, scale, 0.0f,0.0f, &ix0,&iy0,&ix1,&iy1); + + // if empty, return NULL + if (ix0 == ix1 || iy0 == iy1) + return NULL; + + ix0 -= padding; + iy0 -= padding; + ix1 += padding; + iy1 += padding; + + w = (ix1 - ix0); + h = (iy1 - iy0); + + if (width ) *width = w; + if (height) *height = h; + if (xoff ) *xoff = ix0; + if (yoff ) *yoff = iy0; + + // invert for y-downwards bitmaps + scale_y = -scale_y; + + { + int x,y,i,j; + float *precompute; + stbtt_vertex *verts; + int num_verts = stbtt_GetGlyphShape(info, glyph, &verts); + data = (unsigned char *) STBTT_malloc(w * h, info->userdata); + precompute = (float *) STBTT_malloc(num_verts * sizeof(float), info->userdata); + + for (i=0,j=num_verts-1; i < num_verts; j=i++) { + if (verts[i].type == STBTT_vline) { + float x0 = verts[i].x*scale_x, y0 = verts[i].y*scale_y; + float x1 = verts[j].x*scale_x, y1 = verts[j].y*scale_y; + float dist = (float) STBTT_sqrt((x1-x0)*(x1-x0) + (y1-y0)*(y1-y0)); + precompute[i] = (dist == 0) ? 0.0f : 1.0f / dist; + } else if (verts[i].type == STBTT_vcurve) { + float x2 = verts[j].x *scale_x, y2 = verts[j].y *scale_y; + float x1 = verts[i].cx*scale_x, y1 = verts[i].cy*scale_y; + float x0 = verts[i].x *scale_x, y0 = verts[i].y *scale_y; + float bx = x0 - 2*x1 + x2, by = y0 - 2*y1 + y2; + float len2 = bx*bx + by*by; + if (len2 != 0.0f) + precompute[i] = 1.0f / (bx*bx + by*by); + else + precompute[i] = 0.0f; + } else + precompute[i] = 0.0f; + } + + for (y=iy0; y < iy1; ++y) { + for (x=ix0; x < ix1; ++x) { + float val; + float min_dist = 999999.0f; + float sx = (float) x + 0.5f; + float sy = (float) y + 0.5f; + float x_gspace = (sx / scale_x); + float y_gspace = (sy / scale_y); + + int winding = stbtt__compute_crossings_x(x_gspace, y_gspace, num_verts, verts); // @OPTIMIZE: this could just be a rasterization, but needs to be line vs. non-tesselated curves so a new path + + for (i=0; i < num_verts; ++i) { + float x0 = verts[i].x*scale_x, y0 = verts[i].y*scale_y; + + // check against every point here rather than inside line/curve primitives -- @TODO: wrong if multiple 'moves' in a row produce a garbage point, and given culling, probably more efficient to do within line/curve + float dist2 = (x0-sx)*(x0-sx) + (y0-sy)*(y0-sy); + if (dist2 < min_dist*min_dist) + min_dist = (float) STBTT_sqrt(dist2); + + if (verts[i].type == STBTT_vline) { + float x1 = verts[i-1].x*scale_x, y1 = verts[i-1].y*scale_y; + + // coarse culling against bbox + //if (sx > STBTT_min(x0,x1)-min_dist && sx < STBTT_max(x0,x1)+min_dist && + // sy > STBTT_min(y0,y1)-min_dist && sy < STBTT_max(y0,y1)+min_dist) + float dist = (float) STBTT_fabs((x1-x0)*(y0-sy) - (y1-y0)*(x0-sx)) * precompute[i]; + STBTT_assert(i != 0); + if (dist < min_dist) { + // check position along line + // x' = x0 + t*(x1-x0), y' = y0 + t*(y1-y0) + // minimize (x'-sx)*(x'-sx)+(y'-sy)*(y'-sy) + float dx = x1-x0, dy = y1-y0; + float px = x0-sx, py = y0-sy; + // minimize (px+t*dx)^2 + (py+t*dy)^2 = px*px + 2*px*dx*t + t^2*dx*dx + py*py + 2*py*dy*t + t^2*dy*dy + // derivative: 2*px*dx + 2*py*dy + (2*dx*dx+2*dy*dy)*t, set to 0 and solve + float t = -(px*dx + py*dy) / (dx*dx + dy*dy); + if (t >= 0.0f && t <= 1.0f) + min_dist = dist; + } + } else if (verts[i].type == STBTT_vcurve) { + float x2 = verts[i-1].x *scale_x, y2 = verts[i-1].y *scale_y; + float x1 = verts[i ].cx*scale_x, y1 = verts[i ].cy*scale_y; + float box_x0 = STBTT_min(STBTT_min(x0,x1),x2); + float box_y0 = STBTT_min(STBTT_min(y0,y1),y2); + float box_x1 = STBTT_max(STBTT_max(x0,x1),x2); + float box_y1 = STBTT_max(STBTT_max(y0,y1),y2); + // coarse culling against bbox to avoid computing cubic unnecessarily + if (sx > box_x0-min_dist && sx < box_x1+min_dist && sy > box_y0-min_dist && sy < box_y1+min_dist) { + int num=0; + float ax = x1-x0, ay = y1-y0; + float bx = x0 - 2*x1 + x2, by = y0 - 2*y1 + y2; + float mx = x0 - sx, my = y0 - sy; + float res[3],px,py,t,it; + float a_inv = precompute[i]; + if (a_inv == 0.0) { // if a_inv is 0, it's 2nd degree so use quadratic formula + float a = 3*(ax*bx + ay*by); + float b = 2*(ax*ax + ay*ay) + (mx*bx+my*by); + float c = mx*ax+my*ay; + if (a == 0.0) { // if a is 0, it's linear + if (b != 0.0) { + res[num++] = -c/b; + } + } else { + float discriminant = b*b - 4*a*c; + if (discriminant < 0) + num = 0; + else { + float root = (float) STBTT_sqrt(discriminant); + res[0] = (-b - root)/(2*a); + res[1] = (-b + root)/(2*a); + num = 2; // don't bother distinguishing 1-solution case, as code below will still work + } + } + } else { + float b = 3*(ax*bx + ay*by) * a_inv; // could precompute this as it doesn't depend on sample point + float c = (2*(ax*ax + ay*ay) + (mx*bx+my*by)) * a_inv; + float d = (mx*ax+my*ay) * a_inv; + num = stbtt__solve_cubic(b, c, d, res); + } + if (num >= 1 && res[0] >= 0.0f && res[0] <= 1.0f) { + t = res[0], it = 1.0f - t; + px = it*it*x0 + 2*t*it*x1 + t*t*x2; + py = it*it*y0 + 2*t*it*y1 + t*t*y2; + dist2 = (px-sx)*(px-sx) + (py-sy)*(py-sy); + if (dist2 < min_dist * min_dist) + min_dist = (float) STBTT_sqrt(dist2); + } + if (num >= 2 && res[1] >= 0.0f && res[1] <= 1.0f) { + t = res[1], it = 1.0f - t; + px = it*it*x0 + 2*t*it*x1 + t*t*x2; + py = it*it*y0 + 2*t*it*y1 + t*t*y2; + dist2 = (px-sx)*(px-sx) + (py-sy)*(py-sy); + if (dist2 < min_dist * min_dist) + min_dist = (float) STBTT_sqrt(dist2); + } + if (num >= 3 && res[2] >= 0.0f && res[2] <= 1.0f) { + t = res[2], it = 1.0f - t; + px = it*it*x0 + 2*t*it*x1 + t*t*x2; + py = it*it*y0 + 2*t*it*y1 + t*t*y2; + dist2 = (px-sx)*(px-sx) + (py-sy)*(py-sy); + if (dist2 < min_dist * min_dist) + min_dist = (float) STBTT_sqrt(dist2); + } + } + } + } + if (winding == 0) + min_dist = -min_dist; // if outside the shape, value is negative + val = onedge_value + pixel_dist_scale * min_dist; + if (val < 0) + val = 0; + else if (val > 255) + val = 255; + data[(y-iy0)*w+(x-ix0)] = (unsigned char) val; + } + } + STBTT_free(precompute, info->userdata); + STBTT_free(verts, info->userdata); + } + return data; +} + +STBTT_DEF unsigned char * stbtt_GetCodepointSDF(const stbtt_fontinfo *info, float scale, int codepoint, int padding, unsigned char onedge_value, float pixel_dist_scale, int *width, int *height, int *xoff, int *yoff) +{ + return stbtt_GetGlyphSDF(info, scale, stbtt_FindGlyphIndex(info, codepoint), padding, onedge_value, pixel_dist_scale, width, height, xoff, yoff); +} + +STBTT_DEF void stbtt_FreeSDF(unsigned char *bitmap, void *userdata) +{ + STBTT_free(bitmap, userdata); +} + +////////////////////////////////////////////////////////////////////////////// +// +// font name matching -- recommended not to use this +// + +// check if a utf8 string contains a prefix which is the utf16 string; if so return length of matching utf8 string +static stbtt_int32 stbtt__CompareUTF8toUTF16_bigendian_prefix(stbtt_uint8 *s1, stbtt_int32 len1, stbtt_uint8 *s2, stbtt_int32 len2) +{ + stbtt_int32 i=0; + + // convert utf16 to utf8 and compare the results while converting + while (len2) { + stbtt_uint16 ch = s2[0]*256 + s2[1]; + if (ch < 0x80) { + if (i >= len1) return -1; + if (s1[i++] != ch) return -1; + } else if (ch < 0x800) { + if (i+1 >= len1) return -1; + if (s1[i++] != 0xc0 + (ch >> 6)) return -1; + if (s1[i++] != 0x80 + (ch & 0x3f)) return -1; + } else if (ch >= 0xd800 && ch < 0xdc00) { + stbtt_uint32 c; + stbtt_uint16 ch2 = s2[2]*256 + s2[3]; + if (i+3 >= len1) return -1; + c = ((ch - 0xd800) << 10) + (ch2 - 0xdc00) + 0x10000; + if (s1[i++] != 0xf0 + (c >> 18)) return -1; + if (s1[i++] != 0x80 + ((c >> 12) & 0x3f)) return -1; + if (s1[i++] != 0x80 + ((c >> 6) & 0x3f)) return -1; + if (s1[i++] != 0x80 + ((c ) & 0x3f)) return -1; + s2 += 2; // plus another 2 below + len2 -= 2; + } else if (ch >= 0xdc00 && ch < 0xe000) { + return -1; + } else { + if (i+2 >= len1) return -1; + if (s1[i++] != 0xe0 + (ch >> 12)) return -1; + if (s1[i++] != 0x80 + ((ch >> 6) & 0x3f)) return -1; + if (s1[i++] != 0x80 + ((ch ) & 0x3f)) return -1; + } + s2 += 2; + len2 -= 2; + } + return i; +} + +static int stbtt_CompareUTF8toUTF16_bigendian_internal(char *s1, int len1, char *s2, int len2) +{ + return len1 == stbtt__CompareUTF8toUTF16_bigendian_prefix((stbtt_uint8*) s1, len1, (stbtt_uint8*) s2, len2); +} + +// returns results in whatever encoding you request... but note that 2-byte encodings +// will be BIG-ENDIAN... use stbtt_CompareUTF8toUTF16_bigendian() to compare +STBTT_DEF const char *stbtt_GetFontNameStr(const stbtt_fontinfo *font, int *length, int platformID, int encodingID, int languageID, int nameID) +{ + stbtt_int32 i,count,stringOffset; + stbtt_uint8 *fc = font->data; + stbtt_uint32 offset = font->fontstart; + stbtt_uint32 nm = stbtt__find_table(fc, offset, "name"); + if (!nm) return NULL; + + count = ttUSHORT(fc+nm+2); + stringOffset = nm + ttUSHORT(fc+nm+4); + for (i=0; i < count; ++i) { + stbtt_uint32 loc = nm + 6 + 12 * i; + if (platformID == ttUSHORT(fc+loc+0) && encodingID == ttUSHORT(fc+loc+2) + && languageID == ttUSHORT(fc+loc+4) && nameID == ttUSHORT(fc+loc+6)) { + *length = ttUSHORT(fc+loc+8); + return (const char *) (fc+stringOffset+ttUSHORT(fc+loc+10)); + } + } + return NULL; +} + +static int stbtt__matchpair(stbtt_uint8 *fc, stbtt_uint32 nm, stbtt_uint8 *name, stbtt_int32 nlen, stbtt_int32 target_id, stbtt_int32 next_id) +{ + stbtt_int32 i; + stbtt_int32 count = ttUSHORT(fc+nm+2); + stbtt_int32 stringOffset = nm + ttUSHORT(fc+nm+4); + + for (i=0; i < count; ++i) { + stbtt_uint32 loc = nm + 6 + 12 * i; + stbtt_int32 id = ttUSHORT(fc+loc+6); + if (id == target_id) { + // find the encoding + stbtt_int32 platform = ttUSHORT(fc+loc+0), encoding = ttUSHORT(fc+loc+2), language = ttUSHORT(fc+loc+4); + + // is this a Unicode encoding? + if (platform == 0 || (platform == 3 && encoding == 1) || (platform == 3 && encoding == 10)) { + stbtt_int32 slen = ttUSHORT(fc+loc+8); + stbtt_int32 off = ttUSHORT(fc+loc+10); + + // check if there's a prefix match + stbtt_int32 matchlen = stbtt__CompareUTF8toUTF16_bigendian_prefix(name, nlen, fc+stringOffset+off,slen); + if (matchlen >= 0) { + // check for target_id+1 immediately following, with same encoding & language + if (i+1 < count && ttUSHORT(fc+loc+12+6) == next_id && ttUSHORT(fc+loc+12) == platform && ttUSHORT(fc+loc+12+2) == encoding && ttUSHORT(fc+loc+12+4) == language) { + slen = ttUSHORT(fc+loc+12+8); + off = ttUSHORT(fc+loc+12+10); + if (slen == 0) { + if (matchlen == nlen) + return 1; + } else if (matchlen < nlen && name[matchlen] == ' ') { + ++matchlen; + if (stbtt_CompareUTF8toUTF16_bigendian_internal((char*) (name+matchlen), nlen-matchlen, (char*)(fc+stringOffset+off),slen)) + return 1; + } + } else { + // if nothing immediately following + if (matchlen == nlen) + return 1; + } + } + } + + // @TODO handle other encodings + } + } + return 0; +} + +static int stbtt__matches(stbtt_uint8 *fc, stbtt_uint32 offset, stbtt_uint8 *name, stbtt_int32 flags) +{ + stbtt_int32 nlen = (stbtt_int32) STBTT_strlen((char *) name); + stbtt_uint32 nm,hd; + if (!stbtt__isfont(fc+offset)) return 0; + + // check italics/bold/underline flags in macStyle... + if (flags) { + hd = stbtt__find_table(fc, offset, "head"); + if ((ttUSHORT(fc+hd+44) & 7) != (flags & 7)) return 0; + } + + nm = stbtt__find_table(fc, offset, "name"); + if (!nm) return 0; + + if (flags) { + // if we checked the macStyle flags, then just check the family and ignore the subfamily + if (stbtt__matchpair(fc, nm, name, nlen, 16, -1)) return 1; + if (stbtt__matchpair(fc, nm, name, nlen, 1, -1)) return 1; + if (stbtt__matchpair(fc, nm, name, nlen, 3, -1)) return 1; + } else { + if (stbtt__matchpair(fc, nm, name, nlen, 16, 17)) return 1; + if (stbtt__matchpair(fc, nm, name, nlen, 1, 2)) return 1; + if (stbtt__matchpair(fc, nm, name, nlen, 3, -1)) return 1; + } + + return 0; +} + +static int stbtt_FindMatchingFont_internal(unsigned char *font_collection, char *name_utf8, stbtt_int32 flags) +{ + stbtt_int32 i; + for (i=0;;++i) { + stbtt_int32 off = stbtt_GetFontOffsetForIndex(font_collection, i); + if (off < 0) return off; + if (stbtt__matches((stbtt_uint8 *) font_collection, off, (stbtt_uint8*) name_utf8, flags)) + return off; + } +} + +#if defined(__GNUC__) || defined(__clang__) +#pragma GCC diagnostic push +#pragma GCC diagnostic ignored "-Wcast-qual" +#endif + +STBTT_DEF int stbtt_BakeFontBitmap(const unsigned char *data, int offset, + float pixel_height, unsigned char *pixels, int pw, int ph, + int first_char, int num_chars, stbtt_bakedchar *chardata) +{ + return stbtt_BakeFontBitmap_internal((unsigned char *) data, offset, pixel_height, pixels, pw, ph, first_char, num_chars, chardata); +} + +STBTT_DEF int stbtt_GetFontOffsetForIndex(const unsigned char *data, int index) +{ + return stbtt_GetFontOffsetForIndex_internal((unsigned char *) data, index); +} + +STBTT_DEF int stbtt_GetNumberOfFonts(const unsigned char *data) +{ + return stbtt_GetNumberOfFonts_internal((unsigned char *) data); +} + +STBTT_DEF int stbtt_InitFont(stbtt_fontinfo *info, const unsigned char *data, int offset) +{ + return stbtt_InitFont_internal(info, (unsigned char *) data, offset); +} + +STBTT_DEF int stbtt_FindMatchingFont(const unsigned char *fontdata, const char *name, int flags) +{ + return stbtt_FindMatchingFont_internal((unsigned char *) fontdata, (char *) name, flags); +} + +STBTT_DEF int stbtt_CompareUTF8toUTF16_bigendian(const char *s1, int len1, const char *s2, int len2) +{ + return stbtt_CompareUTF8toUTF16_bigendian_internal((char *) s1, len1, (char *) s2, len2); +} + +#if defined(__GNUC__) || defined(__clang__) +#pragma GCC diagnostic pop +#endif + +#endif // STB_TRUETYPE_IMPLEMENTATION + + +// FULL VERSION HISTORY +// +// 1.19 (2018-02-11) OpenType GPOS kerning (horizontal only), STBTT_fmod +// 1.18 (2018-01-29) add missing function +// 1.17 (2017-07-23) make more arguments const; doc fix +// 1.16 (2017-07-12) SDF support +// 1.15 (2017-03-03) make more arguments const +// 1.14 (2017-01-16) num-fonts-in-TTC function +// 1.13 (2017-01-02) support OpenType fonts, certain Apple fonts +// 1.12 (2016-10-25) suppress warnings about casting away const with -Wcast-qual +// 1.11 (2016-04-02) fix unused-variable warning +// 1.10 (2016-04-02) allow user-defined fabs() replacement +// fix memory leak if fontsize=0.0 +// fix warning from duplicate typedef +// 1.09 (2016-01-16) warning fix; avoid crash on outofmem; use alloc userdata for PackFontRanges +// 1.08 (2015-09-13) document stbtt_Rasterize(); fixes for vertical & horizontal edges +// 1.07 (2015-08-01) allow PackFontRanges to accept arrays of sparse codepoints; +// allow PackFontRanges to pack and render in separate phases; +// fix stbtt_GetFontOFfsetForIndex (never worked for non-0 input?); +// fixed an assert() bug in the new rasterizer +// replace assert() with STBTT_assert() in new rasterizer +// 1.06 (2015-07-14) performance improvements (~35% faster on x86 and x64 on test machine) +// also more precise AA rasterizer, except if shapes overlap +// remove need for STBTT_sort +// 1.05 (2015-04-15) fix misplaced definitions for STBTT_STATIC +// 1.04 (2015-04-15) typo in example +// 1.03 (2015-04-12) STBTT_STATIC, fix memory leak in new packing, various fixes +// 1.02 (2014-12-10) fix various warnings & compile issues w/ stb_rect_pack, C++ +// 1.01 (2014-12-08) fix subpixel position when oversampling to exactly match +// non-oversampled; STBTT_POINT_SIZE for packed case only +// 1.00 (2014-12-06) add new PackBegin etc. API, w/ support for oversampling +// 0.99 (2014-09-18) fix multiple bugs with subpixel rendering (ryg) +// 0.9 (2014-08-07) support certain mac/iOS fonts without an MS platformID +// 0.8b (2014-07-07) fix a warning +// 0.8 (2014-05-25) fix a few more warnings +// 0.7 (2013-09-25) bugfix: subpixel glyph bug fixed in 0.5 had come back +// 0.6c (2012-07-24) improve documentation +// 0.6b (2012-07-20) fix a few more warnings +// 0.6 (2012-07-17) fix warnings; added stbtt_ScaleForMappingEmToPixels, +// stbtt_GetFontBoundingBox, stbtt_IsGlyphEmpty +// 0.5 (2011-12-09) bugfixes: +// subpixel glyph renderer computed wrong bounding box +// first vertex of shape can be off-curve (FreeSans) +// 0.4b (2011-12-03) fixed an error in the font baking example +// 0.4 (2011-12-01) kerning, subpixel rendering (tor) +// bugfixes for: +// codepoint-to-glyph conversion using table fmt=12 +// codepoint-to-glyph conversion using table fmt=4 +// stbtt_GetBakedQuad with non-square texture (Zer) +// updated Hello World! sample to use kerning and subpixel +// fixed some warnings +// 0.3 (2009-06-24) cmap fmt=12, compound shapes (MM) +// userdata, malloc-from-userdata, non-zero fill (stb) +// 0.2 (2009-03-11) Fix unsigned/signed char warnings +// 0.1 (2009-03-09) First public release +// + +/* +------------------------------------------------------------------------------ +This software is available under 2 licenses -- choose whichever you prefer. +------------------------------------------------------------------------------ +ALTERNATIVE A - MIT License +Copyright (c) 2017 Sean Barrett +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. +------------------------------------------------------------------------------ +ALTERNATIVE B - Public Domain (www.unlicense.org) +This is free and unencumbered software released into the public domain. +Anyone is free to copy, modify, publish, use, compile, sell, or distribute this +software, either in source code form or as a compiled binary, for any purpose, +commercial or non-commercial, and by any means. +In jurisdictions that recognize copyright laws, the author or authors of this +software dedicate any and all copyright interest in the software to the public +domain. We make this dedication for the benefit of the public at large and to +the detriment of our heirs and successors. We intend this dedication to be an +overt act of relinquishment in perpetuity of all present and future rights to +this software under copyright law. +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 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. +------------------------------------------------------------------------------ +*/ diff --git a/public/blah.h b/public/blah.h new file mode 100644 index 0000000..761685e --- /dev/null +++ b/public/blah.h @@ -0,0 +1,52 @@ +#pragma once + +#include +#include +#include +#include +#include + +#include +#include +#include + +#include +#include +#include + +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include + +#include +#include +#include +#include + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include \ No newline at end of file diff --git a/public/blah/app.cpp b/public/blah/app.cpp new file mode 100644 index 0000000..5b99d21 --- /dev/null +++ b/public/blah/app.cpp @@ -0,0 +1,238 @@ +#include +#include +#include +#include + +#include +#include +#include + +using namespace Blah; + +namespace +{ + static Config app_config; + static bool app_is_running = false; + static bool app_is_exiting = false; +} + +Config::Config() +{ + name = nullptr; + width = 0; + height = 0; + target_framerate = 60; + max_updates = 5; + + graphics = GfxAPI::Any; + on_startup = nullptr; + on_shutdown = nullptr; + on_update = nullptr; + on_render = nullptr; + on_exit_request = App::exit; + on_info = nullptr; + on_warn = nullptr; + on_error = nullptr; +} + +bool App::run(const Config* c) +{ + BLAH_ASSERT(!app_is_running, "The Application is already running"); + BLAH_ASSERT(c != nullptr, "The Application requires a valid Config"); + BLAH_ASSERT(c->name != nullptr, "The Application Name cannot be null"); + BLAH_ASSERT(c->width > 0 && c->height > 0, "The Width and Height must be larget than 0"); + BLAH_ASSERT(c->max_updates > 0, "Max Updates must be >= 1"); + BLAH_ASSERT(c->target_framerate > 0, "Target Framerate must be >= 1"); + + app_config = *c; + app_is_running = true; + app_is_exiting = false; + + Log::print("Starting Up ..."); + + // figure out the graphics api + if (app_config.graphics == GfxAPI::Any) + { + app_config.graphics = Internal::Graphics::pick_api(); + if (app_config.graphics == GfxAPI::Any) + { + Log::error("Failed to find a supported graphics api"); + return false; + } + } + + // initialize the system + if (!Internal::Platform::init(&app_config)) + { + Log::error("Failed to initialize system module"); + return false; + } + + // initialize graphics + if (!Internal::Graphics::init(app_config.graphics)) + { + Log::error("Failed to initialize graphics module"); + return false; + } + + // input + Internal::Input::init(); + + // startup + if (app_config.on_startup != nullptr) + app_config.on_startup(); + + uint64_t time_last = Internal::Platform::time(); + uint64_t time_accumulator = 0; + + // display window + Internal::Platform::ready(); + + while (!app_is_exiting) + { + // poll system events + Internal::Platform::frame(); + + // update at a fixed timerate + { + uint64_t time_target = (uint64_t)((1.0f / app_config.target_framerate) * 1000); + uint64_t time_curr = Internal::Platform::time(); + uint64_t time_diff = time_curr - time_last; + time_last = time_curr; + time_accumulator += time_diff; + + // do not let us run too fast + while (time_accumulator < time_target) + { + Internal::Platform::sleep((int)(time_target - time_accumulator)); + + time_curr = Internal::Platform::time(); + time_diff = time_curr - time_last; + time_last = time_curr; + time_accumulator += time_diff; + } + + // Do not allow us to fall behind too many updates + // (otherwise we'll get spiral of death) + uint64_t time_maximum = app_config.max_updates * time_target; + if (time_accumulator > time_maximum) + time_accumulator = time_maximum; + + // do as many updates as we can + while (time_accumulator >= time_target) + { + time_accumulator -= time_target; + + Time::milliseconds += time_target; + Time::delta = 1.0f / app_config.target_framerate; + Time::previous_elapsed = Time::elapsed; + Time::elapsed += Time::delta; + + Internal::Input::frame(); + Internal::Graphics::frame(); + + Time::pause_timer -= Time::delta; + if (Time::pause_timer <= 0) + { + if (app_config.on_update != nullptr) + app_config.on_update(); + } + } + } + + // render + { + Internal::Graphics::before_render(); + + if (app_config.on_render != nullptr) + app_config.on_render(); + + Internal::Graphics::after_render(); + Internal::Platform::present(); + } + + } + + Log::print("Shutting down ..."); + + // shutdown + if (app_config.on_shutdown != nullptr) + app_config.on_shutdown(); + + Internal::Graphics::shutdown(); + Internal::Platform::shutdown(); + + // clear static state + Log::print("Exited"); + app_is_running = false; + app_is_exiting = false; + Time::milliseconds = 0; + Time::elapsed = 0; + Time::delta = 0; + + return true; +} + +bool App::is_running() +{ + return app_is_running; +} + +void App::exit() +{ + if (!app_is_exiting && app_is_running) + app_is_exiting = true; +} + +const Config* App::config() +{ + return &app_config; +} + +const char* App::path() +{ + return Internal::Platform::app_path(); +} + +const char* App::user_path() +{ + return Internal::Platform::user_path(); +} + +int App::width() +{ + int w, h; + Internal::Platform::get_size(&w, &h); + return w; +} + +int App::height() +{ + int w, h; + Internal::Platform::get_size(&w, &h); + return h; +} + +int App::draw_width() +{ + int w, h; + Internal::Platform::get_draw_size(&w, &h); + return w; +} + +int App::draw_height() +{ + int w, h; + Internal::Platform::get_draw_size(&w, &h); + return h; +} + +float App::content_scale() +{ + return Internal::Platform::get_content_scale(); +} + +void App::fullscreen(bool enabled) +{ + Internal::Platform::set_fullscreen(enabled); +} \ No newline at end of file diff --git a/public/blah/app.h b/public/blah/app.h new file mode 100644 index 0000000..9e5be58 --- /dev/null +++ b/public/blah/app.h @@ -0,0 +1,73 @@ +#pragma once + +namespace Blah +{ + enum class GfxAPI + { + Any = -1, + OpenGL, + Count + }; + + struct Config + { + const char* name; + int width; + int height; + int max_updates; + int target_framerate; + + GfxAPI graphics; + + void (*on_startup)(); + void (*on_shutdown)(); + void (*on_update)(); + void (*on_render)(); + void (*on_exit_request)(); + + void (*on_info)(const char* text); + void (*on_warn)(const char* text); + void (*on_error)(const char* text); + + Config(); + }; + + namespace App + { + // Runs the application + bool run(const Config* config); + + // Returns whether the application is running + bool is_running(); + + // Exits the application + void exit(); + + // Gets the config data used to run the application + const Config* config(); + + // Gets the working path + const char* path(); + + // Gets the user path + const char* user_path(); + + // Gets the width of the window + int width(); + + // Gets the height of the window + int height(); + + // Gets the drawable width of the window + int draw_width(); + + // Gets the drawable height of the window + int draw_height(); + + // Gets the content scale based on the OS + float content_scale(); + + // Toggles fullscreen + void fullscreen(bool enabled); + } +} \ No newline at end of file diff --git a/public/blah/containers/list.h b/public/blah/containers/list.h new file mode 100644 index 0000000..b87b98b --- /dev/null +++ b/public/blah/containers/list.h @@ -0,0 +1,314 @@ +#pragma once +#include +#include +#include +#include + +namespace Blah +{ + template + class List + { + private: + // internal list buffer + T* m_buffer; + + // total elements + int m_count; + + // total capacity + int m_capacity; + + public: + + List(); + List(int capacity); + List(const List& src); + List(List&& src) noexcept; + ~List(); + + List& operator=(const List& src); + List& operator=(List&& src) noexcept; + + // reserves the given amount of capacity + void reserve(int new_capacity); + + // adds an element to the list + void add(const T& item); + + template + void emplace(Args&&...args); + + // moves an element into the list + void add(T&& item); + + // expands the list by the given amount, and returns a pointer to the first element + T* expand(int amount = 1); + + // returns a reference to the element at the given index + T& operator[](int index); + + // returns a reference to the element at the given index + const T& operator[](int index) const; + + // returns a pointer to the first element + T* begin() { return m_buffer; } + + // returns a pointer to the first element + const T* begin() const { return m_buffer; } + + // returns a pointer to the last element + T* end() { return m_buffer + m_count; } + + // returns a pointer to the last element + const T* end() const { return m_buffer + m_count; } + + // clears all elements + void clear(); + + // removes the element at the index + void remove_at(int index); + + // checks whether the given value is in the List (uses == operator) + bool contains(const T& item) const; + + // returns the index of the given value in the list (uses == operator, -1 if not in list) + int index_of(const T& item) const; + + // pops the top element off the list + T pop(); + + // returns the total number of elements + int count() const { return m_count; } + + // returns the internal buffer capacity of the list + int capacity() const { return m_capacity; } + + // removes all elements and disposes the internal buffer + void dispose(); + }; + + template + List::List() + { + m_buffer = nullptr; + m_count = m_capacity = 0; + } + + template + List::List(int capacity) + { + m_buffer = nullptr; + m_count = m_capacity = 0; + Reserve(m_capacity); + } + + template + List::List(const List& src) + { + m_buffer = (T*)(::operator new (sizeof(T) * src.m_capacity)); + m_count = src.m_count; + m_capacity = src.m_capacity; + + for (int n = 0; n < m_count; n++) + new (m_buffer + n) T(*(src.m_buffer + n)); + } + + template + List::List(List&& src) noexcept + { + m_buffer = src.m_buffer; + m_count = src.m_count; + m_capacity = src.m_capacity; + src.m_buffer = nullptr; + src.m_count = src.m_capacity = 0; + } + + template + List& List::operator=(const List& src) + { + clear(); + + if (m_capacity < src.m_count) + reserve(src.m_count); + m_count = src.m_count; + + for (int n = 0; n < m_count; n++) + new (m_buffer + n) T(*(src.m_buffer + n)); + + return *this; + } + + template + List& List::operator=(List&& src) noexcept + { + dispose(); + m_buffer = src.m_buffer; + m_count = src.m_count; + m_capacity = src.m_capacity; + src.m_buffer = nullptr; + src.m_count = src.m_capacity = 0; + return *this; + } + + template + List::~List() + { + dispose(); + } + + template + void List::reserve(int new_capacity) + { + if (new_capacity >= m_capacity) + { + auto last_capacity = m_capacity; + + if (m_capacity <= 0) + m_capacity = 8; + + while (new_capacity >= m_capacity) + m_capacity = m_capacity * 2; + + T* new_buffer = (T*)::operator new (sizeof(T) * m_capacity); + + if (std::is_trivially_copyable::value) + { + memcpy(new_buffer, m_buffer, sizeof(T) * m_count); + } + else + { + for (int n = 0; n < m_count; n++) + new (new_buffer + n) T(std::move(m_buffer[n])); + } + + ::operator delete (m_buffer, sizeof(T) * last_capacity); + m_buffer = new_buffer; + } + } + + template + T* List::expand(int amount) + { + reserve(m_count + amount); + for (int n = m_count; n < m_count + amount; n++) + new (m_buffer + n) T(); + m_count += amount; + return (m_buffer + m_count - amount); + } + + template + void List::add(const T& item) + { + reserve(m_count + 1); + new (m_buffer + m_count) T(item); + m_count++; + } + + template + template + void List::emplace(Args&&...args) + { + reserve(m_count + 1); + new (m_buffer + m_count) T(args); + m_count++; + } + + template + void List::add(T&& item) + { + reserve(m_count + 1); + new (m_buffer + m_count) T(std::move(item)); + m_count++; + } + + template + T& List::operator[](int index) + { + BLAH_ASSERT(index >= 0 && index < m_count, "Index is out of range"); + return m_buffer[index]; + } + + template + const T& List::operator[](int index) const + { + BLAH_ASSERT(index >= 0 && index < m_count, "Index is out of range"); + return m_buffer[index]; + } + + template + void List::clear() + { + for (T* it = m_buffer; it < m_buffer + m_count; it++) + it->~T(); + m_count = 0; + } + + template + void List::remove_at(int index) + { + BLAH_ASSERT(index >= 0 && index < m_count, "Index is out of range"); + + m_buffer[index].~T(); + + if (index < m_count - 1) + { + int diff = (m_count - index - 1); + if (diff <= 0) + diff = 0; + + if (std::is_trivially_copyable::value) + { + memmove(m_buffer + index, m_buffer + index + 1, (size_t)diff * sizeof(T)); + } + else + { + for (int i = index; i < m_count - 1; i++) + m_buffer[i] = std::move(m_buffer[i + 1]); + m_buffer[m_count - 1].~T(); + } + } + + m_count--; + } + + template + inline bool List::contains(const T& item) const + { + for (int i = 0; i < m_count; i++) + if (*(m_buffer + i) == item) + return true; + return false; + } + + template + inline int List::index_of(const T& item) const + { + for (int i = 0; i < m_count; i++) + if (*(m_buffer + i) == item) + return i; + return -1; + } + + template + T List::pop() + { + BLAH_ASSERT(m_count > 0, "There are no elements to pop"); + m_count--; + + T item = m_buffer[m_count]; + m_buffer[m_count].~T(); + return item; + } + + template + void List::dispose() + { + clear(); + ::operator delete (m_buffer, sizeof(T) * m_capacity); + m_buffer = nullptr; + m_count = 0; + m_capacity = 0; + } + +} diff --git a/public/blah/containers/stacklist.h b/public/blah/containers/stacklist.h new file mode 100644 index 0000000..aad44cc --- /dev/null +++ b/public/blah/containers/stacklist.h @@ -0,0 +1,212 @@ +#pragma once +#include + +namespace Blah +{ + template + class StackList + { + private: + // internal list buffer + T m_buffer[Size]; + + // total elements + int m_count; + + public: + + StackList(); + StackList(const StackList& src); + StackList(StackList&& src) noexcept; + + StackList& operator=(const StackList& src); + StackList& operator=(StackList&& src) noexcept; + + // adds an element to the list + void add(const T& item); + + // moves an element into the list + void add(T&& item); + + // expands the list by the given amount, and returns a pointer to the first element + T* expand(int amount = 1); + + // returns a reference to the element at the given index + T& operator[](int index); + + // returns a reference to the element at the given index + const T& operator[](int index) const; + + // returns a pointer to the first element + T* begin() { return m_buffer; } + + // returns a pointer to the first element + const T* begin() const { return m_buffer; } + + // returns a pointer to the last element + T* end() { return m_buffer + m_count; } + + // returns a pointer to the last element + const T* end() const { return m_buffer + m_count; } + + // clears all elements + void clear(); + + // removes the element at the index + void remove_at(int index); + + // checks whether the given value is in the StackList (uses == operator) + bool contains(const T& item) const; + + // returns the index of the given value in the list (uses == operator, -1 if not in list) + int index_of(const T& item) const; + + // pops the top element off the list + T& pop(); + + // returns the total number of elements + int count() const { return m_count; } + + // returns the internal buffer capacity of the list + int capacity() const { return Size; } + }; + + template + StackList::StackList() + { + m_count = 0; + } + + template + StackList::StackList(const StackList& src) + { + m_count = src.m_count; + + for (int n = 0; n < m_count; n++) + m_buffer[n] = src.m_buffer[n]; + } + + template + StackList::StackList(StackList&& src) noexcept + { + m_count = src.m_count; + for (int n = 0; n < m_count; n++) + m_buffer[n] = std::move(src.m_buffer[n]); + src.m_count = 0; + } + + template + StackList& StackList::operator=(const StackList& src) + { + clear(); + + m_count = src.m_count; + for (int n = 0; n < m_count; n++) + m_buffer[n] = src.m_buffer[n]; + + return *this; + } + + template + StackList& StackList::operator=(StackList&& src) noexcept + { + clear(); + + m_count = src.m_count; + for (int n = 0; n < m_count; n++) + m_buffer[n] = std::move(src.m_buffer[n]); + src.m_count = 0; + + return *this; + } + + template + T* StackList::expand(int amount) + { + BLAH_ASSERT(m_count + amount < Size, "Exceeded StackList capacity"); + + for (int n = m_count; n < m_count + amount; n++) + m_buffer[n] = T(); + m_count += amount; + + return (m_buffer + m_count - amount); + } + + template + void StackList::add(const T& item) + { + BLAH_ASSERT(m_count < Size, "Exceeded StackList capacity"); + m_buffer[m_count] = item; + m_count++; + } + + template + void StackList::add(T&& item) + { + BLAH_ASSERT(m_count < Size, "Exceeded StackList capacity"); + m_buffer[m_count] = std::move(item); + m_count++; + } + + template + T& StackList::operator[](int index) + { + BLAH_ASSERT(index >= 0 && index < m_count, "Index is out of range"); + return m_buffer[index]; + } + + template + const T& StackList::operator[](int index) const + { + BLAH_ASSERT(index >= 0 && index < m_count, "Index is out of range"); + return m_buffer[index]; + } + + template + void StackList::clear() + { + m_count = 0; + } + + template + void StackList::remove_at(int index) + { + BLAH_ASSERT(index >= 0 && index < m_count, "Index is out of range"); + + if (index < m_count - 1) + { + for (int n = index; n < m_count - 1; n++) + m_buffer[n] = m_buffer[n + 1]; + m_buffer[m_count] = T(); + } + + m_count--; + } + + template + bool StackList::contains(const T& item) const + { + for (int i = 0; i < m_count; i++) + if (*(m_buffer + i) == item) + return true; + return false; + } + + template + int StackList::index_of(const T& item) const + { + for (int i = 0; i < m_count; i++) + if (*(m_buffer + i) == item) + return i; + return -1; + } + + template + T& StackList::pop() + { + BLAH_ASSERT(m_count > 0, "There are no elements to pop"); + m_count--; + return m_buffer[m_count]; + } + +} diff --git a/public/blah/containers/str.cpp b/public/blah/containers/str.cpp new file mode 100644 index 0000000..0358fac --- /dev/null +++ b/public/blah/containers/str.cpp @@ -0,0 +1,385 @@ +#include +#include // for strcpy etc +#include // for format methods +#include // for sprintf +#include // toupper + +using namespace Blah; + +const uint16_t UTF8_LEAD_SURROGATE_MIN = 0xd800u; +const uint16_t UTF8_LEAD_SURROGATE_MAX = 0xdbffu; +const uint16_t UTF8_TRAIL_SURROGATE_MIN = 0xdc00u; +const uint16_t UTF8_TRAIL_SURROGATE_MAX = 0xdfffu; +const uint16_t UTF8_LEAD_OFFSET = UTF8_LEAD_SURROGATE_MIN - (0x10000 >> 10); +const uint32_t UTF8_SURROGATE_OFFSET = 0x10000u - (UTF8_LEAD_SURROGATE_MIN << 10) - UTF8_TRAIL_SURROGATE_MIN; + +char Str::empty_buffer[1] = { '\0' }; + +bool Str::operator==(const Str& rhs) const +{ + return strcmp(cstr(), rhs.cstr()) == 0; +} + +bool Str::operator!=(const Str& rhs) const +{ + return strcmp(cstr(), rhs.cstr()) != 0; +} + +bool Str::operator==(const char* rhs) const +{ + return strcmp(cstr(), rhs) == 0; +} + +bool Str::operator!=(const char* rhs) const +{ + return strcmp(cstr(), rhs) != 0; +} + +void Str::reserve(int size) +{ + int buffer_length = size + 1; + if (buffer_length > m_capacity) + { + if (m_capacity <= 0) + m_capacity = 16; + + while (m_capacity < buffer_length) + m_capacity *= 2; + + // expand from local buffer + if (m_buffer == nullptr) + { + char* local = data(); + m_buffer = new char[m_capacity]; + strncpy(m_buffer, local, m_local_size); + m_buffer[m_local_size] = '\0'; + } + // expand from empty buffer + else if (m_buffer == empty_buffer) + { + m_buffer = new char[m_capacity]; + m_buffer[0] = '\0'; + } + // expand from existing heap buffer + else + { + char* new_buffer = new char[m_capacity]; + memcpy(new_buffer, m_buffer, m_length); + new_buffer[m_length] = '\0'; + delete[] m_buffer; + m_buffer = new_buffer; + } + } +} + +void Str::set_length(int len) +{ + reserve(len); + data()[len] = '\0'; + m_length = len; +} + +Str& Str::append(char c) +{ + reserve(m_length + 1); + data()[m_length++] = c; + data()[m_length] = '\0'; + return *this; +} + +Str& Str::append(uint32_t c) +{ + // one octet + if (c < 0x80) + { + append((char)c); + } + // two octets + else if (c < 0x800) + { + append((char)((c >> 6) | 0xc0)); + append((char)((c & 0x3f) | 0x80)); + } + // three octets + else if (c < 0x10000) + { + append((char)((c >> 12) | 0xe0)); + append((char)(((c >> 6) & 0x3f) | 0x80)); + append((char)((c & 0x3f) | 0x80)); + } + // four octets + else + { + append((char)((c >> 18) | 0xf0)); + append((char)(((c >> 12) & 0x3f) | 0x80)); + append((char)(((c >> 6) & 0x3f) | 0x80)); + append((char)((c & 0x3f) | 0x80)); + } + + return *this; +} + +Str& Str::append(const char* start, const char* end) +{ + if (end == nullptr) + end = start + strlen(start); + + int add = (int)(end - start); + if (add <= 0) + return *this; + + reserve(m_length + add); + memcpy(data() + m_length, start, add); + m_length += add; + data()[m_length] = '\0'; + + return *this; +} + +Str& Str::append(const Str& str, int start, int end) +{ + if (end < 0) end = str.m_length; + if (end > str.m_length) end = str.m_length; + if (start < 0) start = 0; + if (start > end) start = end; + + return append(str.begin() + start, str.begin() + end); +} + +Str& Str::append_fmt(const char* fmt, ...) +{ + int add, diff; + + // determine arg m_length + va_list args; + va_start(args, fmt); + add = vsnprintf(NULL, 0, fmt, args); + va_end(args); + + // reserve + reserve(m_length + add); + diff = m_capacity - m_length; + if (diff <= 0) diff = 0; + + // print out + va_start(args, fmt); + vsnprintf(data() + m_length, (size_t)diff, fmt, args); + va_end(args); + + // increment size + m_length += add; + data()[m_length] = '\0'; + + return *this; +} + +Str& Str::append_utf16(const uint16_t* start, const uint16_t* end, bool swapEndian) +{ + while (start != end) + { + uint16_t next = (*start++); + if (swapEndian) + next = ((next & 0xff) << 8 | ((next & 0xff00) >> 8)); + + uint32_t cp = 0xffff & next; + + if ((cp >= UTF8_LEAD_SURROGATE_MIN && cp <= UTF8_LEAD_SURROGATE_MAX)) + { + next = (*start++); + if (swapEndian) + next = ((next & 0xff) << 8 | ((next & 0xff00) >> 8)); + + uint32_t trail = 0xffff & next; + cp = (cp << 10) + trail + UTF8_SURROGATE_OFFSET; + } + + append(cp); + } + + return *this; +} + +Str& Str::trim() +{ + const char* s = begin(); + const char* e = end() - 1; + + while (isspace(*s) && s != e) + s++; + while (isspace(*e) && s != e) + e--; + + set(s, e + 1); + return *this; +} + +bool Str::starts_with(const char* str, bool ignoreCase) const +{ + int len = (int)strlen(str); + if (len > m_length || len <= 0) + return false; + + const char* a = data(); + const char* b = str; + + if (ignoreCase) + { + for (int n = 0; n < len; n++) + if (tolower(a[n]) != tolower(b[n])) + return false; + } + else + { + for (int n = 0; n < len; n++) + if (a[n] != b[n]) + return false; + } + + return true; +} + +bool Str::contains(const char* str, bool ignoreCase) const +{ + int len = (int)strlen(str); + if (len > m_length || len <= 0) + return false; + + const char* a = data(); + const char* b = str; + + for (int start = 0; start < m_length - len + 1; start++) + { + bool match = true; + + if (ignoreCase) + { + for (int n = 0; n < len && match; n++) + if (tolower(a[start + n]) != tolower(b[n])) + match = false; + } + else + { + for (int n = 0; n < len && match; n++) + if (a[start + n] != b[n]) + match = false; + } + + if (match) + return true; + } + + return false; +} + +bool Str::ends_with(const char* str, bool ignoreCase) const +{ + int len = (int)strlen(str); + if (len > m_length || len <= 0) + return false; + + const char* a = data(); + const char* b = str; + + if (ignoreCase) + { + for (int n = m_length - len, i = 0; n < m_length; n++, i++) + if (tolower(a[n]) != tolower(b[i])) + return false; + } + else + { + for (int n = m_length - len, i = 0; n < m_length; n++, i++) + if (a[n] != b[i]) + return false; + } + + return true; +} + +int Str::first_index_of(char ch) const +{ + const char* ptr = data(); + for (int n = 0; n < m_length; n++) + if (ptr[n] == ch) + return n; + return -1; +} + +int Str::last_index_of(char ch) const +{ + const char* ptr = data(); + for (int n = m_length - 1; n >= 0; n--) + if (ptr[n] == ch) + return n; + return -1; +} + +Str Str::substr(int start) const +{ + if (start < 0) start = 0; + if (start > m_length) start = m_length; + + return Str(data() + start); +} + +Str Str::substr(int start, int end) const +{ + if (start < 0) start = 0; + if (start > m_length) start = m_length; + if (end < 0) end = m_length + end; + if (end < start) end = start; + if (end > m_length) end = m_length; + + return Str(data() + start, data() + end); +} + +void Str::replace(char c, char r) +{ + char* ptr = data(); + for (int n = 0; n < m_length; n++) + if (ptr[n] == c) + ptr[n] = r; +} + +void Str::clear() +{ + if (m_capacity > 0) + data()[0] = '\0'; + m_length = 0; +} + +void Str::dispose() +{ + if (m_buffer != nullptr && m_buffer != empty_buffer) + delete[] m_buffer; + + if (m_local_size > 0) + { + m_buffer = nullptr; + m_capacity = m_local_size; + data()[0] = '\0'; + } + else + { + m_buffer = empty_buffer; + m_capacity = 0; + } + + m_length = 0; +} + +void Str::set(const char* start, const char* end) +{ + // find the end + if (end == nullptr) + end = start + strlen(start); + + // reserve + m_length = (int)(end - start); + reserve(m_length); + + // copy the data over + char* ptr = data(); + memcpy(ptr, start, m_length); + ptr[m_length] = '\0'; +} \ No newline at end of file diff --git a/public/blah/containers/str.h b/public/blah/containers/str.h new file mode 100644 index 0000000..60a7ea4 --- /dev/null +++ b/public/blah/containers/str.h @@ -0,0 +1,185 @@ +#pragma once +#include + +namespace Blah +{ + class Str + { + public: + Str() { m_buffer = empty_buffer; m_length = m_capacity = m_local_size = 0; } + Str(const char* start, const char* end = nullptr) : Str() { set(start, end); } + Str(const Str& str) : Str() { set(str); } + + char& operator[](int index) { return data()[index]; } + const char& operator[](int index) const { return data()[index]; } + + // equality operators + bool operator==(const Str& rhs) const; + bool operator!=(const Str& rhs) const; + bool operator==(const char* rhs) const; + bool operator!=(const char* rhs) const; + + // assignment operator + Str& operator=(const Str& rhs) { set(rhs.cstr(), rhs.cstr() + rhs.m_length); return *this; } + Str& operator=(const char* rhs) { set(rhs, nullptr); return *this; } + + // append string + Str& operator+=(const Str& rhs) { return append(rhs); } + + // append cstr + Str& operator+=(const char* rhs) { return append(rhs); } + + // combine string + Str operator+(const Str& rhs) { Str str; str.append(*this).append(rhs); return str; } + + // combine cstr + Str operator+(const char* rhs) { Str str; str.append(*this).append(rhs); return str; } + + // implicit cast to cstr + operator char* () { return cstr(); } + + // implicit cast to cstr + operator const char* () const { return cstr(); } + + // returns a pointer to the null-terminated string buffer + char* cstr() { return data(); } + + // returns a pointer to the null-terminated string buffer + const char* cstr() const { return data(); } + + // returns a pointer to the start of the buffer + const char* begin() const { return data(); } + + // returns a pointer to the end of the buffer + const char* end() const { return data() + m_length; } + + // returns the length of the string + int length() const { return m_length; } + + // returns the capacity of the string + int capacity() const { return m_capacity; } + + // returns the capacity of the string's stack buffer + int stack_capacity() const { return m_local_size; } + + // sets the length of the string. + // this does not set the value of the string! + void set_length(int length); + + // ensures the string has the given capacity + void reserve(int capacity); + + // appends the given character + Str& append(char c); + + // appends the given unicode character + Str& append(uint32_t c); + + // appends the given c string + Str& append(const char* start, const char* end = nullptr); + + // appends the given string + Str& append(const Str& str, int start = 0, int end = -1); + + // appends the given formatted string + Str& append_fmt(const char* fmt, ...); + + // appends a utf16 string + Str& append_utf16(const uint16_t* start, const uint16_t* end = nullptr, bool swapEndian = false); + + // trims whitespace + Str& trim(); + + // returns true if the string begins with the given string + bool starts_with(const Str& str, bool ignoreCase = false) const { return starts_with(str.cstr(), ignoreCase); } + + // returns true if the string begins with the given string + bool starts_with(const char* str, bool ignoreCase = false) const; + + // returns true if the string contains with the given string + bool contains(const Str& str, bool ignoreCase = false) const { return contains(str.cstr(), ignoreCase); } + + // returns true if the string contains with the given string + bool contains(const char* str, bool ignoreCase = false) const; + + // returns true if the string ends with the given string + bool ends_with(const Str& str, bool ignoreCase = false) const { return ends_with(str.cstr(), ignoreCase); } + + // returns true if the string ends with the given string + bool ends_with(const char* str, bool ignoreCase = false) const; + + // returns the first index of the given character, or -1 if it isn't found + int first_index_of(char ch) const; + + // returns the last index of the given character, or -1 if it isn't found + int last_index_of(char ch) const; + + // returns a substring of the string + Str substr(int start) const; + + // returns a substring of the string + Str substr(int start, int end) const; + + // replaces all occurances of the given character in the string + void replace(char c, char r); + + // checks if the string has a length of 0 + bool empty() const { return m_length <= 0; } + + // clears the string length to 0 + void clear(); + + // clears and disposes the internal string buffer + void dispose(); + + ~Str() + { + if (m_buffer != nullptr && m_buffer != empty_buffer) + delete[] m_buffer; + } + + protected: + Str(int local_size) + { + m_buffer = nullptr; + m_length = 0; + m_capacity = local_size; + m_local_size = local_size; + } + + // returns a pointer to the heap buffer or to our stack allocation + char* data() { return (m_buffer != nullptr ? m_buffer : ((char*)(this) + sizeof(Str))); } + const char* data() const { return (m_buffer != nullptr ? m_buffer : ((char*)(this) + sizeof(Str))); } + + void set(const Str& str) { set(str.cstr(), str.cstr() + str.m_length); } + + void set(const char* start, const char* end = nullptr); + + private: + static char empty_buffer[1]; + char* m_buffer; + int m_length; + int m_capacity; + int m_local_size; + }; + + template + class StrOf : public Str + { + private: + char m_local_buffer[T]; + + public: + StrOf() : Str(T) { m_local_buffer[0] = '\0'; } + StrOf(const char* rhs, const char* end = nullptr) : Str(T) { m_local_buffer[0] = '\0'; set(rhs, end); } + StrOf(const Str& rhs) : Str(T) { m_local_buffer[0] = '\0'; set(rhs); } + StrOf(const StrOf& rhs) : Str(T) { m_local_buffer[0] = '\0'; set(rhs); } + + // assignment operators + StrOf& operator=(const char* rhs) { set(rhs); return *this; } + StrOf& operator=(const Str& rhs) { set(rhs); return *this; } + StrOf& operator=(const StrOf& rhs) { set(rhs); return *this; } + }; + + using String = StrOf<64>; +} \ No newline at end of file diff --git a/public/blah/drawing/batch.cpp b/public/blah/drawing/batch.cpp new file mode 100644 index 0000000..5e8c27d --- /dev/null +++ b/public/blah/drawing/batch.cpp @@ -0,0 +1,832 @@ +#include +#include +#include +#include +#include +#include +#include + +using namespace Blah; + +// TODO: +// the line drawings methods aren't consistent +// ex. some draw outside the shape, some draw within it, etc. +// -> make them all draw within the provided shape + +// TODO: +// This needs to be graphics API agnostic + +const ShaderData data = { + // vertex shader + "#version 330\n" + "uniform mat4 u_matrix;\n" + "layout(location=0) in vec2 a_position;\n" + "layout(location=1) in vec2 a_tex;\n" + "layout(location=2) in vec4 a_color;\n" + "layout(location=3) in vec3 a_type;\n" + "out vec2 v_tex;\n" + "out vec4 v_col;\n" + "out vec3 v_type;\n" + "void main(void)\n" + "{\n" + " gl_Position = u_matrix * vec4(a_position.xy, 0, 1);\n" + " v_tex = a_tex;\n" + " v_col = a_color;\n" + " v_type = a_type;\n" + "}", + + // fragment shader + "#version 330\n" + "uniform sampler2D u_texture;\n" + "in vec2 v_tex;\n" + "in vec4 v_col;\n" + "in vec3 v_type;\n" + "out vec4 o_color;\n" + "void main(void)\n" + "{\n" + " vec4 color = texture(u_texture, v_tex);\n" + " o_color = \n" + " v_type.x * color * v_col + \n" + " v_type.y * color.a * v_col + \n" + " v_type.z * v_col;\n" + "}" +}; + +const VertexAttribute attributes[4] = { + { 0, VertexSemantics::Position, VertexAttributeType::Float, 2, false }, + { 1, VertexSemantics::Texcoord0, VertexAttributeType::Float, 2, false }, + { 2, VertexSemantics::Color0, VertexAttributeType::Byte, 4, true }, + { 3, VertexSemantics::Texcoord1, VertexAttributeType::Byte, 3, true }, +}; + +#define MAKE_VERTEX(vert, mat, px, py, tx, ty, c, m, w, f) \ + (vert)->pos.x = ((px) * mat.m11) + ((py) * mat.m21) + mat.m31; \ + (vert)->pos.y = ((px) * mat.m12) + ((py) * mat.m22) + mat.m32; \ + (vert)->tex.x = tx; \ + if (m_batch.flip_vertically) \ + (vert)->tex.y = 1.0f - ty; \ + else \ + (vert)->tex.y = ty; \ + (vert)->col = c; \ + (vert)->mult = m; \ + (vert)->wash = w; \ + (vert)->fill = f; + +#define PUSH_QUAD(px0, py0, px1, py1, px2, py2, px3, py3, tx0, ty0, tx1, ty1, tx2, ty2, tx3, ty3, col0, col1, col2, col3, mult, fill, wash) \ + { \ + m_batch.elements += 2; \ + int* _i = m_indices.expand(6); \ + *_i++ = m_vertices.count() + 0; \ + *_i++ = m_vertices.count() + 1; \ + *_i++ = m_vertices.count() + 2; \ + *_i++ = m_vertices.count() + 0; \ + *_i++ = m_vertices.count() + 2; \ + *_i++ = m_vertices.count() + 3; \ + Vertex* _v = m_vertices.expand(4); \ + MAKE_VERTEX(_v, m_matrix, px0, py0, tx0, ty0, col0, mult, fill, wash); _v++; \ + MAKE_VERTEX(_v, m_matrix, px1, py1, tx1, ty1, col1, mult, fill, wash); _v++; \ + MAKE_VERTEX(_v, m_matrix, px2, py2, tx2, ty2, col2, mult, fill, wash); _v++; \ + MAKE_VERTEX(_v, m_matrix, px3, py3, tx3, ty3, col3, mult, fill, wash); \ + } + +#define PUSH_TRIANGLE(px0, py0, px1, py1, px2, py2, tx0, ty0, tx1, ty1, tx2, ty2, col0, col1, col2, mult, fill, wash) \ + { \ + m_batch.elements += 1; \ + int* _i = m_indices.expand(3); \ + *_i++ = m_vertices.count() + 0; \ + *_i++ = m_vertices.count() + 1; \ + *_i++ = m_vertices.count() + 2; \ + Vertex* _v = m_vertices.expand(3); \ + MAKE_VERTEX(_v, m_matrix, px0, py0, tx0, ty0, col0, mult, fill, wash); _v++; \ + MAKE_VERTEX(_v, m_matrix, px1, py1, tx1, ty1, col1, mult, fill, wash); _v++; \ + MAKE_VERTEX(_v, m_matrix, px2, py2, tx2, ty2, col2, mult, fill, wash); \ + } + +// Compares a Batcher variable, and starts a new batch if it has changed +#define SET_BATCH_VAR(variable) \ + if (m_batch.elements > 0 && variable != m_batch.variable) \ + { \ + m_batches.add(m_batch); \ + m_batch.offset += m_batch.elements; \ + m_batch.elements = 0; \ + } \ + m_batch.variable = variable; + +ShaderRef Batch::m_default_shader; + +Batch::Batch() +{ + matrix_uniform = "u_matrix"; + texture_uniform = "u_texture"; + + clear(); +} + +Batch::~Batch() +{ + dispose(); +} + +void Batch::push_matrix(const Mat3x2& matrix) +{ + m_matrix_stack.add(m_matrix); + m_matrix = matrix * m_matrix; +} + +void Batch::pop_matrix() +{ + m_matrix = m_matrix_stack.pop(); +} + +void Batch::push_scissor(const Rect& scissor) +{ + m_scissor_stack.add(m_batch.scissor); + SET_BATCH_VAR(scissor); +} + +void Batch::pop_scissor() +{ + Rect scissor = m_scissor_stack.pop(); + SET_BATCH_VAR(scissor); +} + +void Batch::push_blend(const BlendMode& blend) +{ + m_blend_stack.add(m_batch.blend); + SET_BATCH_VAR(blend); +} + +void Batch::pop_blend() +{ + BlendMode blend = m_blend_stack.pop(); + SET_BATCH_VAR(blend); +} + +void Batch::push_material(const MaterialRef& material) +{ + m_material_stack.add(m_batch.material); + SET_BATCH_VAR(material); +} + +void Batch::pop_material() +{ + MaterialRef material = m_material_stack.pop(); + SET_BATCH_VAR(material); +} + +void Batch::push_layer(int layer) +{ + m_layer_stack.add(m_batch.layer); + SET_BATCH_VAR(layer); +} + +void Batch::pop_layer() +{ + int layer = m_layer_stack.pop(); + SET_BATCH_VAR(layer); +} + +void Batch::push_color_mode(ColorMode mode) +{ + m_color_mode_stack.add(m_color_mode); + m_color_mode = mode; + + m_tex_mult = (m_color_mode == ColorMode::Normal ? 255 : 0); + m_tex_wash = (m_color_mode == ColorMode::Wash ? 255 : 0); +} + +void Batch::pop_color_mode() +{ + m_color_mode = m_color_mode_stack.pop(); + m_tex_mult = (m_color_mode == ColorMode::Normal ? 255 : 0); + m_tex_wash = (m_color_mode == ColorMode::Wash ? 255 : 0); +} + +void Batch::set_texture(const TextureRef& texture) +{ + if (m_batch.elements > 0 && texture != m_batch.texture && m_batch.texture && m_batch.texture->is_valid()) + { + m_batches.add(m_batch); + m_batch.offset += m_batch.elements; + m_batch.elements = 0; + } + + if (m_batch.texture != texture) + { + m_batch.texture = texture; + m_batch.flip_vertically = Graphics::info()->origin_bottom_left && texture && texture->is_valid() && texture->is_framebuffer(); + } +} + +void Batch::render(const FrameBufferRef& target) +{ + Point size; + if (!target || !target->is_valid()) + size = Point(App::draw_width(), App::draw_height()); + else + size = Point(target->width(), target->height()); + + render(target, Mat4x4::CreateOrthographicOffCenter(0, (float)size.x, (float)size.y, 0, 0.01f, 1000.0f)); +} + +void Batch::render(const FrameBufferRef& target, const Mat4x4& matrix) +{ + // nothing to draw + if ((m_batches.count() <= 0 && m_batch.elements <= 0) || m_indices.count() <= 0) + return; + + // define defaults + { + if (!m_mesh || !m_mesh->is_valid()) + { + m_mesh = Graphics::create_mesh(); + m_mesh->vertex_format(attributes, 4, sizeof(Vertex)); + } + + if (!m_default_shader) + m_default_shader = Graphics::create_shader(&data); + + if (!m_default_material) + m_default_material = Graphics::create_material(m_default_shader); + } + + // upload data + m_mesh->index_data(m_indices.begin(), m_indices.count()); + m_mesh->vertex_data(m_vertices.begin(), m_vertices.count()); + + RenderCall call; + call.target = target; + call.mesh = m_mesh; + call.has_viewport = false; + call.viewport = Rect(); + call.instance_count = 0; + call.depth = Compare::None; + call.cull = Cull::None; + + for (auto& b : m_batches) + render_single_batch(call, b, matrix); + + if (m_batch.elements > 0) + render_single_batch(call, m_batch, matrix); +} + +void Batch::render_single_batch(RenderCall& call, const DrawBatch& b, const Mat4x4& matrix) +{ + call.material = b.material; + if (!call.material || !call.material->is_valid()) + call.material = m_default_material; + + call.material->set_texture(texture_uniform, b.texture, 0); + call.material->set_value(matrix_uniform, &matrix.m11, 16); + + call.blend = b.blend; + call.has_scissor = b.scissor.w >= 0 && b.scissor.h >= 0; + call.scissor = b.scissor; + call.index_start = (int64_t)b.offset * 3; + call.index_count = (int64_t)b.elements * 3; + + Graphics::render(call); +} + +void Batch::clear() +{ + m_matrix = Mat3x2::identity; + m_color_mode = ColorMode::Normal; + m_tex_mult = 255; + m_tex_wash = 0; + + m_vertices.clear(); + m_indices.clear(); + + m_batch.layer = 0; + m_batch.elements = 0; + m_batch.offset = 0; + m_batch.blend = BlendMode::Normal; + m_batch.material.reset(); + m_batch.texture.reset(); + m_batch.scissor.w = m_batch.scissor.h = -1; + m_batch.flip_vertically = false; + + m_matrix_stack.clear(); + m_scissor_stack.clear(); + m_blend_stack.clear(); + m_material_stack.clear(); + m_color_mode_stack.clear(); + m_layer_stack.clear(); + m_batches.clear(); +} + +void Batch::dispose() +{ + m_matrix_stack.dispose(); + m_scissor_stack.dispose(); + m_blend_stack.dispose(); + m_material_stack.dispose(); + m_color_mode_stack.dispose(); + m_layer_stack.dispose(); + m_batches.dispose(); + m_default_material.reset(); + m_mesh.reset(); + m_vertices.dispose(); + m_indices.dispose(); +} + +void Batch::line(const Vec2& from, const Vec2& to, float t, Color color) +{ + line(from, to, t, color, color); +} + +void Batch::line(const Vec2& from, const Vec2& to, float t, Color fromColor, Color toColor) +{ + if (from.x == to.x && from.y == to.y) + return; + + Vec2 normal = (to - from).normal(); + Vec2 perp = Vec2(normal.y, -normal.x); + Vec2 pos0 = from + perp * t * 0.5f; + Vec2 pos1 = to + perp * t * 0.5f; + Vec2 pos2 = to - perp * t * 0.5f; + Vec2 pos3 = from - perp * t * 0.5f; + + PUSH_QUAD( + pos0.x, pos0.y, pos1.x, pos1.y, pos2.x, pos2.y, pos3.x, pos3.y, + 0, 0, 0, 0, 0, 0, 0, 0, + fromColor, toColor, toColor, fromColor, + 0, 0, 255); +} + +void Batch::bezier_line(const Vec2& from, const Vec2& b, const Vec2& to, int steps, float t, Color color) +{ + Vec2 prev = from; + float add = 1.0f / steps; + + for (int i = 1; i < steps; i++) + { + Vec2 at = Vec2::bezier_lerp(from, b, to, add * i); + line(prev, at, t, color); + prev = at; + } + + line(prev, to, t, color); +} + +void Batch::bezier_line(const Vec2& from, const Vec2& b, const Vec2& c, const Vec2& to, int steps, float t, Color color) +{ + Vec2 prev = from; + float add = 1.0f / steps; + + for (int i = 1; i < steps; i++) + { + Vec2 at = Vec2::bezier_lerp(from, b, c, to, add * i); + line(prev, at, t, color); + prev = at; + } + + line(prev, to, t, color); +} + +void Batch::tri(const Vec2& pos0, const Vec2& pos1, const Vec2& pos2, Color color) +{ + PUSH_TRIANGLE( + pos0.x, pos0.y, pos1.x, pos1.y, pos2.x, pos2.y, + 0, 0, 0, 0, 0, 0, + color, color, color, + 0, 0, 255); +} + +void Batch::tri(const Vec2& pos0, const Vec2& pos1, const Vec2& pos2, Color col0, Color col1, Color col2) +{ + PUSH_TRIANGLE( + pos0.x, pos0.y, pos1.x, pos1.y, pos2.x, pos2.y, + 0, 0, 0, 0, 0, 0, + col0, col1, col2, + 0, 0, 255); +} + +void Batch::tri(const Vec2& pos0, const Vec2& pos1, const Vec2& pos2, const Vec2& tex0, const Vec2& tex1, const Vec2& tex2, Color color) +{ + PUSH_TRIANGLE( + pos0.x, pos0.y, pos1.x, pos1.y, pos2.x, pos2.y, + tex0.x, tex0.y, tex1.x, tex1.y, tex2.x, tex2.y, + color, color, color, + m_tex_mult, m_tex_wash, 0); +} + +void Batch::tri(const Vec2& pos0, const Vec2& pos1, const Vec2& pos2, const Vec2& tex0, const Vec2& tex1, const Vec2& tex2, Color col0, Color col1, Color col2) +{ + PUSH_TRIANGLE( + pos0.x, pos0.y, pos1.x, pos1.y, pos2.x, pos2.y, + tex0.x, tex0.y, tex1.x, tex1.y, tex2.x, tex2.y, + col0, col1, col2, + m_tex_mult, m_tex_wash, 0); +} + +void Batch::tri_line(const Vec2& a, const Vec2& b, const Vec2& c, float t, Color color) +{ + +} + +void Batch::rect(const Rect& rect, Color color) +{ + PUSH_QUAD( + rect.x, rect.y, + rect.x + rect.w, rect.y, + rect.x + rect.w, rect.y + rect.h, + rect.x, rect.y + rect.h, + 0, 0, 0, 0, 0, 0, 0, 0, + color, color, color, color, + 0, 0, 255); +} + +void Batch::rect_line(const Rect& rect, float t, Color color) +{ + PUSH_QUAD( + rect.x, rect.y, + rect.x + rect.w - t, rect.y, + rect.x + rect.w - t, rect.y + t, + rect.x, rect.y + t, + 0, 0, 0, 0, 0, 0, 0, 0, + color, color, color, color, + 0, 0, 255); + + PUSH_QUAD( + rect.x + rect.w - t, rect.y, + rect.x + rect.w, rect.y, + rect.x + rect.w, rect.y + rect.h - t, + rect.x + rect.w - t, rect.y + rect.h - t, + 0, 0, 0, 0, 0, 0, 0, 0, + color, color, color, color, + 0, 0, 255); + + PUSH_QUAD( + rect.x + t, rect.y + rect.h - t, + rect.x + rect.w, rect.y + rect.h - t, + rect.x + rect.w, rect.y + rect.h, + rect.x, rect.y + rect.h, + 0, 0, 0, 0, 0, 0, 0, 0, + color, color, color, color, + 0, 0, 255); + + PUSH_QUAD( + rect.x, rect.y + t, + rect.x + t, rect.y + t, + rect.x + t, rect.y + rect.h, + rect.x, rect.y + rect.h, + 0, 0, 0, 0, 0, 0, 0, 0, + color, color, color, color, + 0, 0, 255); +} + +void Batch::rect_rounded(const Rect& rect, float radius, int steps, Color color) +{ + rect_rounded(rect, radius, steps, radius, steps, radius, steps, radius, steps, color); +} + +void Batch::rect_rounded(const Rect& rect, float rtl, int rtl_steps, float rtr, int rtr_steps, float rbr, int rbr_steps, float rbl, int rbl_steps, Color color) +{ + // clamp + rtl = Calc::min(Calc::min(Calc::max(0.0f, rtl), rect.w / 2.0f), rect.h / 2.0f); + rtr = Calc::min(Calc::min(Calc::max(0.0f, rtr), rect.w / 2.0f), rect.h / 2.0f); + rbr = Calc::min(Calc::min(Calc::max(0.0f, rbr), rect.w / 2.0f), rect.h / 2.0f); + rbl = Calc::min(Calc::min(Calc::max(0.0f, rbl), rect.w / 2.0f), rect.h / 2.0f); + + if (rtl <= 0 && rtr <= 0 && rbr <= 0 && rbl <= 0) + { + this->rect(rect, color); + } + else + { + // get corners + Rect tl = Rect(rect.top_left(), Vec2(rtl, rtl)); + Rect tr = Rect(rect.top_right() + Vec2(-rtr, 0), Vec2(rtr, rtr)); + Rect bl = Rect(rect.bottom_left() + Vec2(0, -rbl), Vec2(rbl, rbl)); + Rect br = Rect(rect.bottom_right() + Vec2(-rbr, -rbr), Vec2(rbr, rbr)); + + // rounded corners + semi_circle(tl.bottom_right(), Calc::UP, Calc::LEFT, rtl, rtl_steps, color); + semi_circle(tr.bottom_left(), Calc::UP, Calc::RIGHT, rtr, rtr_steps, color); + semi_circle(bl.top_right(), Calc::DOWN, Calc::LEFT, rbl, rbl_steps, color); + semi_circle(br.top_left(), Calc::DOWN, Calc::RIGHT, rbr, rbr_steps, color); + + // center quads + quad(tl.top_right(), tr.top_left(), tr.bottom_left(), tl.bottom_right(), color); + quad(tr.bottom_left(), tr.bottom_right(), br.top_right(), br.top_left(), color); + quad(bl.top_right(), br.top_left(), br.bottom_left(), bl.bottom_right(), color); + quad(tl.bottom_left(), tl.bottom_right(), bl.top_right(), bl.top_left(), color); + quad(tl.bottom_right(), tr.bottom_left(), br.top_left(), bl.top_right(), color); + } +} + +void Batch::rect_rounded_line(const Rect & rect, float radius, int steps, float t, Color color) +{ + rect_rounded_line(rect, radius, steps, radius, steps, radius, steps, radius, steps, t, color); +} + +void Batch::rect_rounded_line(const Rect& rect, float rtl, int rtl_steps, float rtr, int rtr_steps, float rbr, int rbr_steps, float rbl, int rbl_steps, float t, Color color) +{ + Rect r = rect; + r.x += t * 0.5f; + r.y += t * 0.5f; + r.w -= t; + r.h -= t; + + // clamp + rtl = Calc::min(Calc::min(Calc::max(0.0f, rtl), r.w / 2.0f), r.h / 2.0f); + rtr = Calc::min(Calc::min(Calc::max(0.0f, rtr), r.w / 2.0f), r.h / 2.0f); + rbr = Calc::min(Calc::min(Calc::max(0.0f, rbr), r.w / 2.0f), r.h / 2.0f); + rbl = Calc::min(Calc::min(Calc::max(0.0f, rbl), r.w / 2.0f), r.h / 2.0f); + + if (rtl <= 0 && rtr <= 0 && rbr <= 0 && rbl <= 0) + { + this->rect_line(r, t, color); + } + else + { + // get corners + Rect tl = Rect(r.top_left(), Vec2(rtl, rtl)); + Rect tr = Rect(r.top_right() + Vec2(-rtr, 0), Vec2(rtr, rtr)); + Rect bl = Rect(r.bottom_left() + Vec2(0, -rbl), Vec2(rbl, rbl)); + Rect br = Rect(r.bottom_right() + Vec2(-rbr, -rbr), Vec2(rbr, rbr)); + + // rounded corners + semi_circle_line(tl.bottom_right(), Calc::UP, Calc::LEFT, rtl, rtl_steps, t, color); + semi_circle_line(tr.bottom_left(), Calc::UP, Calc::RIGHT, rtr, rtr_steps, t, color); + semi_circle_line(bl.top_right(), Calc::DOWN, Calc::LEFT, rbl, rbl_steps, t, color); + semi_circle_line(br.top_left(), Calc::DOWN, Calc::RIGHT, rbr, rbr_steps, t, color); + + // connect sides that aren't touching + if (tl.bottom() < bl.top()) + line(tl.bottom_left(), bl.top_left(), t, color); + if (tl.right() < tr.left()) + line(tl.top_right(), tr.top_left(), t, color); + if (tr.bottom() < br.top()) + line(tr.bottom_right(), br.top_right(), t, color); + if (bl.right() < br.left()) + line(bl.bottom_right(), br.bottom_left(), t, color); + } +} + +void Batch::semi_circle(Vec2 center, float start_radians, float end_radians, float radius, int steps, Color centerColor, Color edgeColor) +{ + Vec2 last = Vec2::from_angle(start_radians, radius); + float add = Calc::angle_diff(start_radians, end_radians); + + for (int i = 1; i <= steps; i++) + { + Vec2 next = Vec2::from_angle(start_radians + add * (i / (float)steps), radius); + tri(center + last, center + next, center, edgeColor, edgeColor, centerColor); + last = next; + } +} + +void Batch::semi_circle(Vec2 center, float start_radians, float end_radians, float radius, int steps, Color color) +{ + semi_circle(center, start_radians, end_radians, radius, steps, color, color); +} + +void Batch::semi_circle_line(Vec2 center, float start_radians, float end_radians, float radius, int steps, float t, Color color) +{ + float add = Calc::angle_diff(start_radians, end_radians); + + Vec2 last_inner = Vec2::from_angle(start_radians, radius - t / 2); + Vec2 last_outer = Vec2::from_angle(start_radians, radius + t / 2); + + for (int i = 1; i <= steps; i++) + { + Vec2 next_inner = Vec2::from_angle(start_radians + add * (i / (float)steps), radius - t / 2); + Vec2 next_outer = Vec2::from_angle(start_radians + add * (i / (float)steps), radius + t / 2); + + quad(center + last_inner, center + last_outer, center + next_outer, center + next_inner, color); + + last_inner = next_inner; + last_outer = next_outer; + } +} + +void Batch::circle(const Vec2 center, float radius, int steps, Color color) +{ + Vec2 last = Vec2(radius, 0); + + for (int i = 1; i <= steps; i++) + { + float radians = (i / (float)steps) * Calc::TAU; + Vec2 next = Vec2::from_angle(radians, radius); + tri(center + last, center + next, center, color); + last = next; + } +} + +void Batch::circle_line(const Vec2 center, float radius, float t, int steps, Color color) +{ + radius -= t * 0.5f; + Vec2 last = Vec2(radius, 0); + + for (int i = 1; i <= steps; i++) + { + float radians = (i / (float)steps) * Calc::TAU; + Vec2 next = Vec2(Calc::cos(radians), Calc::sin(radians)) * radius; + line(center + last, center + next, t, color); + last = next; + } +} + +void Batch::quad(const Vec2& pos0, const Vec2& pos1, const Vec2& pos2, const Vec2& pos3, Color color) +{ + PUSH_QUAD( + pos0.x, pos0.y, pos1.x, pos1.y, pos2.x, pos2.y, pos3.x, pos3.y, + 0, 0, 0, 0, 0, 0, 0, 0, + color, color, color, color, + 0, 0, 255); +} + +void Batch::quad(const Vec2& pos0, const Vec2& pos1, const Vec2& pos2, const Vec2& pos3, Color col0, Color col1, Color col2, Color col3) +{ + PUSH_QUAD( + pos0.x, pos0.y, pos1.x, pos1.y, pos2.x, pos2.y, pos3.x, pos3.y, + 0, 0, 0, 0, 0, 0, 0, 0, + col0, col1, col2, col3, + 0, 0, 255); +} + +void Batch::quad(const Vec2& pos0, const Vec2& pos1, const Vec2& pos2, const Vec2& pos3, const Vec2& tex0, const Vec2& tex1, const Vec2& tex2, const Vec2& tex3, Color color) +{ + PUSH_QUAD( + pos0.x, pos0.y, pos1.x, pos1.y, pos2.x, pos2.y, pos3.x, pos3.y, + tex0.x, tex0.y, tex1.x, tex1.y, tex2.x, tex2.y, tex3.x, tex3.y, + color, color, color, color, + m_tex_mult, m_tex_wash, 0); +} + +void Batch::quad(const Vec2& pos0, const Vec2& pos1, const Vec2& pos2, const Vec2& pos3, const Vec2& tex0, const Vec2& tex1, const Vec2& tex2, const Vec2& tex3, Color col0, Color col1, Color col2, Color col3) +{ + PUSH_QUAD( + pos0.x, pos0.y, pos1.x, pos1.y, pos2.x, pos2.y, pos3.x, pos3.y, + tex0.x, tex0.y, tex1.x, tex1.y, tex2.x, tex2.y, tex3.x, tex3.y, + col0, col1, col2, col3, + m_tex_mult, m_tex_wash, 0); +} + +void Batch::quad_line(const Vec2& a, const Vec2& b, const Vec2& c, const Vec2& d, float t, Color color) +{ + line(a, b, t, color); + line(b, c, t, color); + line(c, d, t, color); + line(d, a, t, color); +} + +void Batch::arrow_head(const Vec2& point_pos, float radians, float side_len, Color color) +{ + arrow_head(point_pos, point_pos - Vec2::from_angle(radians), side_len, color); +} + +void Batch::arrow_head(const Vec2& point_pos, const Vec2& from_pos, float side_len, Color color) +{ + float height = sqrtf(side_len * side_len - powf(side_len / 2, 2)); + Vec2 dir = (point_pos - from_pos).normal(); + Vec2 perp = dir.perpendicular(); + Vec2 base = point_pos - dir * height; + + tri(point_pos, base + perp * side_len / 2, base - perp * side_len / 2, color); +} + +void Batch::tex() +{ + +} + +void Batch::tex(const Subtexture& sub, const Vec2& pos, Color color) +{ + if (!sub.texture || !sub.texture->is_valid()) + { + PUSH_QUAD( + pos.x + sub.draw_coords[0].x, pos.y + sub.draw_coords[0].y, + pos.x + sub.draw_coords[1].x, pos.y + sub.draw_coords[1].y, + pos.x + sub.draw_coords[2].x, pos.y + sub.draw_coords[2].y, + pos.x + sub.draw_coords[3].x, pos.y + sub.draw_coords[3].y, + 0, 0, 0, 0, 0, 0, 0, 0, + color, color, color, color, + 0, 0, 255); + } + else + { + set_texture(sub.texture); + + PUSH_QUAD( + pos.x + sub.draw_coords[0].x, pos.y + sub.draw_coords[0].y, + pos.x + sub.draw_coords[1].x, pos.y + sub.draw_coords[1].y, + pos.x + sub.draw_coords[2].x, pos.y + sub.draw_coords[2].y, + pos.x + sub.draw_coords[3].x, pos.y + sub.draw_coords[3].y, + sub.tex_coords[0].x, sub.tex_coords[0].y, + sub.tex_coords[1].x, sub.tex_coords[1].y, + sub.tex_coords[2].x, sub.tex_coords[2].y, + sub.tex_coords[3].x, sub.tex_coords[3].y, + color, color, color, color, + m_tex_mult, m_tex_wash, 0); + } +} + +void Batch::tex(const Subtexture& sub, const Vec2& pos, const Vec2& origin, const Vec2& scale, float rotation, Color color) +{ + push_matrix(Mat3x2::create_transform(pos, origin, scale, rotation)); + + if (!sub.texture || !sub.texture->is_valid()) + { + PUSH_QUAD( + sub.draw_coords[0].x, sub.draw_coords[0].y, + sub.draw_coords[1].x, sub.draw_coords[1].y, + sub.draw_coords[2].x, sub.draw_coords[2].y, + sub.draw_coords[3].x, sub.draw_coords[3].y, + 0, 0, 0, 0, 0, 0, 0, 0, + color, color, color, color, + 0, 0, 255); + } + else + { + set_texture(sub.texture); + + PUSH_QUAD( + sub.draw_coords[0].x, sub.draw_coords[0].y, + sub.draw_coords[1].x, sub.draw_coords[1].y, + sub.draw_coords[2].x, sub.draw_coords[2].y, + sub.draw_coords[3].x, sub.draw_coords[3].y, + sub.tex_coords[0].x, sub.tex_coords[0].y, + sub.tex_coords[1].x, sub.tex_coords[1].y, + sub.tex_coords[2].x, sub.tex_coords[2].y, + sub.tex_coords[3].x, sub.tex_coords[3].y, + color, color, color, color, + m_tex_mult, m_tex_wash, 0); + } + + pop_matrix(); +} + +void Batch::str(const SpriteFont& font, const String& text, const Vec2& pos, Color color) +{ + str(font, text, pos, TextAlign::TopLeft, font.size(), color); +} + +void Batch::str(const SpriteFont& font, const String& text, const Vec2& pos, TextAlign align, float size, Color color) +{ + push_matrix( + Mat3x2::create_scale(size / font.size()) * + Mat3x2::create_translation(pos) + ); + + Vec2 offset; + + if ((align & TextAlign::Left) == TextAlign::Left) + offset.x = 0; + else if ((align & TextAlign::Right) == TextAlign::Right) + offset.x -= font.width_of_line(text); + else + offset.x -= font.width_of_line(text) * 0.5f; + + if ((align & TextAlign::Top) == TextAlign::Top) + offset.y = font.ascent() + font.descent(); + else if ((align & TextAlign::Bottom) == TextAlign::Bottom) + offset.y = font.ascent() + font.descent() - font.height_of(text); + else + offset.y = font.ascent() + font.descent() - font.height_of(text) * 0.5f; + + for (int i = 0, l = text.length(); i < l; i++) + { + if (text[i] == '\n') + { + // increment y + offset.y += font.line_height(); + + // re-align X for this line + if ((align & TextAlign::Left) == TextAlign::Left) + offset.x = 0; + else if ((align & TextAlign::Right) == TextAlign::Right) + offset.x = -font.width_of_line(text, i + 1); + else + offset.x = -font.width_of_line(text, i + 1) * 0.5f; + + continue; + } + + // TODO: + // This doesn't parse Unicode! + // It will assume it's a 1-byte ASCII char which is incorrect + auto ch = font[text[i]]; + + if (!ch.subtexture.texture || !ch.subtexture.texture->is_valid()) + { + Vec2 at = offset + ch.offset; + + if (i > 0 && text[i - 1] != '\n') + { + // TODO: + // This doesn't parse Unicode! + at.x += font.get_kerning(text[i], text[i - 1]); + } + + tex(ch.subtexture, at, color); + } + + offset.x += ch.advance; + } + + pop_matrix(); +} \ No newline at end of file diff --git a/public/blah/drawing/batch.h b/public/blah/drawing/batch.h new file mode 100644 index 0000000..c2380c0 --- /dev/null +++ b/public/blah/drawing/batch.h @@ -0,0 +1,169 @@ +#pragma once +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +namespace Blah +{ + + enum class ColorMode + { + Normal, + Wash + }; + + enum class TextAlign : char + { + Center = 0, + Left = 1 << 1, + Right = 1 << 2, + Top = 1 << 3, + Bottom = 1 << 4, + + TopLeft = Top | Left, + TopRight = Top | Right, + BottomLeft = Bottom | Left, + BottomRight = Bottom | Right + }; + + inline TextAlign operator|(TextAlign lhs, TextAlign rhs) { return static_cast(static_cast(lhs) | static_cast(rhs)); } + inline TextAlign operator&(TextAlign lhs, TextAlign rhs) { return static_cast(static_cast(lhs) & static_cast(rhs)); } + + class Batch + { + public: + Batch(); + Batch(const Batch& other) = delete; + Batch& operator=(const Batch& other) = delete; + ~Batch(); + + void push_matrix(const Mat3x2& matrix); + void pop_matrix(); + void push_scissor(const Rect& scissor); + void pop_scissor(); + void push_blend(const BlendMode& blend); + void pop_blend(); + void push_material(const MaterialRef& material); + void pop_material(); + void push_layer(int layer); + void pop_layer(); + void push_color_mode(ColorMode mode); + void pop_color_mode(); + + void set_texture(const TextureRef& texture); + + void render(const FrameBufferRef& target); + void render(const FrameBufferRef& target, const Mat4x4& matrix); + + void clear(); + void dispose(); + + void line(const Vec2& from, const Vec2& to, float t, Color color); + void line(const Vec2& from, const Vec2& to, float t, Color fromColor, Color toColor); + + void bezier_line(const Vec2& from, const Vec2& b, const Vec2& to, int steps, float t, Color color); + void bezier_line(const Vec2& from, const Vec2& b, const Vec2& c, const Vec2& to, int steps, float t, Color color); + + void tri(const Vec2& pos0, const Vec2& pos1, const Vec2& pos2, Color color); + void tri(const Vec2& pos0, const Vec2& pos1, const Vec2& pos2, Color col0, Color col1, Color col2); + void tri(const Vec2& pos0, const Vec2& pos1, const Vec2& pos2, const Vec2& tex0, const Vec2& tex1, const Vec2& tex2, Color color); + void tri(const Vec2& pos0, const Vec2& pos1, const Vec2& pos2, const Vec2& tex0, const Vec2& tex1, const Vec2& tex2, Color col0, Color col1, Color col2); + + void tri_line(const Vec2& a, const Vec2& b, const Vec2& c, float t, Color color); + + void rect(const Rect& rect, Color color); + void rect_line(const Rect& rect, float t, Color color); + void rect_rounded(const Rect& rect, float radius, int steps, Color color); + void rect_rounded(const Rect& rect, float rtl, int rtl_steps, float rtr, int rtr_steps, float rbr, int rbr_steps, float rbl, int rbl_steps, Color color); + void rect_rounded_line(const Rect& rect, float radius, int steps, float t, Color color); + void rect_rounded_line(const Rect& rect, float rtl, int rtl_steps, float rtr, int rtr_steps, float rbr, int rbr_steps, float rbl, int rbl_steps, float t, Color color); + + void semi_circle(Vec2 center, float start_radians, float end_radians, float radius, int steps, Color centerColor, Color edgeColor); + void semi_circle(Vec2 center, float start_radians, float end_radians, float radius, int steps, Color color); + void semi_circle_line(Vec2 center, float start_radians, float end_radians, float radius, int steps, float t, Color color); + + void circle(const Vec2 center, float radius, int steps, Color color); + void circle_line(const Vec2 center, float raidus, float t, int steps, Color color); + + void quad(const Vec2& pos0, const Vec2& pos1, const Vec2& pos2, const Vec2& pos3, Color color); + void quad(const Vec2& pos0, const Vec2& pos1, const Vec2& pos2, const Vec2& pos3, Color col0, Color col1, Color col2, Color col3); + void quad(const Vec2& pos0, const Vec2& pos1, const Vec2& pos2, const Vec2& pos3, const Vec2& tex0, const Vec2& tex1, const Vec2& tex2, const Vec2& tex3, Color color); + void quad(const Vec2& pos0, const Vec2& pos1, const Vec2& pos2, const Vec2& pos3, const Vec2& tex0, const Vec2& tex1, const Vec2& tex2, const Vec2& tex3, Color col0, Color col1, Color col2, Color col3); + void quad_line(const Vec2& a, const Vec2& b, const Vec2& c, const Vec2& d, float t, Color color); + + void arrow_head(const Vec2& point_pos, float radians, float side_len, Color color); + void arrow_head(const Vec2& point_pos, const Vec2& from_pos, float side_len, Color color); + + void tex(); + void tex(const Subtexture& subtexture, const Vec2& pos, Color color); + void tex(const Subtexture& subtexture, const Vec2& pos, const Vec2& origin, const Vec2& scale, float rotation, Color color); + void str(const SpriteFont& font, const String& text, const Vec2& pos, Color color); + void str(const SpriteFont& font, const String& text, const Vec2& pos, TextAlign align, float size, Color color); + + const char* matrix_uniform; + const char* texture_uniform; + + private: + + struct Vertex + { + Vec2 pos; + Vec2 tex; + Color col; + + uint8_t mult; + uint8_t wash; + uint8_t fill; + + static VertexAttribute attributes[6]; + }; + + struct DrawBatch + { + int layer; + int offset; + int elements; + MaterialRef material; + BlendMode blend; + TextureRef texture; + bool flip_vertically; + Rect scissor; + + DrawBatch() : + layer(0), + offset(0), + elements(0), + blend(BlendMode::Normal), + flip_vertically(false), + scissor(0, 0, -1, -1) {} + }; + + static ShaderRef m_default_shader; + MaterialRef m_default_material; + MeshRef m_mesh; + Mat3x2 m_matrix; + ColorMode m_color_mode; + uint8_t m_tex_mult; + uint8_t m_tex_wash; + DrawBatch m_batch; + List m_vertices; + List m_indices; + + List m_matrix_stack; + List m_scissor_stack; + List m_blend_stack; + List m_material_stack; + List m_color_mode_stack; + List m_layer_stack; + List m_batches; + + void render_single_batch(RenderCall& call, const DrawBatch& b, const Mat4x4& matrix); + }; +} \ No newline at end of file diff --git a/public/blah/drawing/spritefont.cpp b/public/blah/drawing/spritefont.cpp new file mode 100644 index 0000000..aae9bc7 --- /dev/null +++ b/public/blah/drawing/spritefont.cpp @@ -0,0 +1,250 @@ +#include +#include +#include + +using namespace Blah; + +SpriteFont::SpriteFont() +{ + m_size = 0; + m_ascent = 0; + m_descent = 0; + m_line_gap = 0; +} + +const uint32_t ascii[]{ 32, 128, 0 }; +const uint32_t* SpriteFont::ASCII = ascii; + +SpriteFont::SpriteFont(const char* file, float size) +{ + build(file, size, ASCII); +} + +SpriteFont::SpriteFont(const char* file, float size, const uint32_t* charset) +{ + build(file, size, charset); +} + +SpriteFont::SpriteFont(const Font& font, float size) +{ + build(font, size, ASCII); +} + +SpriteFont::SpriteFont(const Font& font, float size, const uint32_t* charset) +{ + build(font, size, charset); +} + +SpriteFont::SpriteFont(SpriteFont&& src) noexcept +{ + m_name = src.m_name; + m_size = src.m_size; + m_ascent = src.m_ascent; + m_descent = src.m_descent; + m_line_gap = src.m_line_gap; + m_characters = std::move(src.m_characters); + m_kerning = std::move(src.m_kerning); + m_atlas = std::move(src.m_atlas); +} + +SpriteFont::~SpriteFont() +{ + dispose(); +} + +void SpriteFont::dispose() +{ + m_atlas.dispose(); + m_characters.clear(); + m_kerning.clear(); + m_name.dispose(); +} + +SpriteFont& SpriteFont::operator=(SpriteFont && src) noexcept +{ + m_name = src.m_name; + m_size = src.m_size; + m_ascent = src.m_ascent; + m_descent = src.m_descent; + m_line_gap = src.m_line_gap; + m_characters = std::move(src.m_characters); + m_kerning = std::move(src.m_kerning); + m_atlas = std::move(src.m_atlas); + return *this; +} + +float SpriteFont::width_of(const String& text) const +{ + float width = 0; + float lineWidth = 0; + for (auto it = text.begin(); it != text.end(); it++) + { + if (*it == '\n') + lineWidth = 0; + + // TODO: this doesn't account for Unicode values! + uint32_t codepoint = *it; + + lineWidth += this->operator[](codepoint).advance; + if (lineWidth > width) + width = lineWidth; + } + + return width; +} + +float SpriteFont::width_of_line(const String& text, int start) const +{ + if (start < 0) return 0; + if (start >= text.length()) return 0; + + float lineWidth = 0; + for (auto it = text.begin() + start; it != text.end(); it++) + { + if (*it == '\n') + return lineWidth; + + // TODO: this doesn't account for Unicode values! + uint32_t codepoint = *it; + + lineWidth += this->operator[](codepoint).advance; + } + + return lineWidth; +} + +float SpriteFont::height_of(const String& text) const +{ + if (text.length() <= 0) + return 0; + + float height = line_height(); + for (auto it = text.begin(); it != text.end(); it++) + { + if (*it == '\n') + height += line_height(); + } + + return height - m_line_gap; +} + +void SpriteFont::build(const char* file, float sz, const uint32_t* charset) +{ + dispose(); + + Font font(file); + if (font.IsValid()) + build(font, sz, charset); +} + +void SpriteFont::build(const Font& font, float size, const uint32_t* charset) +{ + dispose(); + + float scale = font.GetScale(size); + + m_name = font.FamilyName(); + m_ascent = font.Ascent() * scale; + m_descent = font.Descent() * scale; + m_line_gap = font.LineGap() * scale; + m_size = size; + + Packer packer; + packer.spacing = 0; + packer.padding = 1; + packer.max_size = 8192; + packer.power_of_two = true; + + std::map glyphs; + List buffer; + + auto ranges = charset; + while (*ranges != 0) + { + auto from = *ranges; + auto to = *(ranges + 1); + + BLAH_ASSERT(to >= from, "Charset Range must be in pairs of [min,max]"); + + for (auto i = from; i < to; i++) + { + auto glyph = font.GetGlyph(i); + if (glyph <= 0) + continue; + + glyphs[i] = glyph; + + // add character + Font::Char ch = font.GetCharacter(glyph, scale); + m_characters[i].advance = ch.advance; + m_characters[i].offset = Vec2(ch.offsetX, ch.offsetY); + + // pack glyph + if (ch.hasGlyph) + { + if (buffer.count() < ch.width * ch.height) + buffer.expand(ch.width * ch.height - buffer.count()); + + if (font.GetBitmap(ch, buffer.begin())) + packer.add(i, ch.width, ch.height, buffer.begin()); + } + } + + ranges += 2; + } + + buffer.dispose(); + packer.pack(); + + for (auto& it : packer.pages) + m_atlas.add(Graphics::create_texture(it)); + + // add character subtextures + for (auto& it : packer.entries) + if (!it.empty) + m_characters[(uint32_t)it.id].subtexture = Subtexture(m_atlas[it.page], it.packed, it.frame); + + // add kerning + for (auto a = glyphs.begin(); a != glyphs.end(); a++) + for (auto b = glyphs.begin(); b != glyphs.end(); b++) + set_kerning(a->first, b->first, font.GetKerning(a->second, b->second, scale)); +} + +float SpriteFont::get_kerning(uint32_t codepoint0, uint32_t codepoint1) const +{ + Tuple index; + index.first = codepoint0; + index.second = codepoint1; + + auto it = m_kerning.find(index); + if (it != m_kerning.end()) + return it->second; + return 0.0f; +} + +void SpriteFont::set_kerning(uint32_t codepoint0, uint32_t codepoint1, float value) +{ + Tuple index; + index.first = codepoint0; + index.second = codepoint1; + m_kerning[index] = value; +} + + +const SpriteFont::Character& SpriteFont::get_character(uint32_t codepoint) const +{ + static const Character empty; + auto it = m_characters.find(codepoint); + if (it != m_characters.end()) + return it->second; + return empty; +} + +const SpriteFont::Character& SpriteFont::operator[](uint32_t codepoint) const +{ + static const Character empty; + auto it = m_characters.find(codepoint); + if (it != m_characters.end()) + return it->second; + return empty; +} \ No newline at end of file diff --git a/public/blah/drawing/spritefont.h b/public/blah/drawing/spritefont.h new file mode 100644 index 0000000..ade510e --- /dev/null +++ b/public/blah/drawing/spritefont.h @@ -0,0 +1,93 @@ +#pragma once +#include +#include +#include +#include +#include +#include + +namespace Blah +{ + class Font; + + class SpriteFont + { + public: + struct Character + { + Subtexture subtexture; + float advance = 0; + Vec2 offset; + }; + private: + + // properties + String m_name; + float m_size; + float m_ascent; + float m_descent; + float m_line_gap; + + // tuple between two characters + struct Tuple { uint32_t first; uint32_t second; }; + struct TupleCompare + { + bool operator() (const Tuple& lhs, const Tuple& rhs) const + { + return lhs.first < rhs.first && lhs.second < rhs.second; + } + }; + + // charset & kerning maps + std::map m_characters; + std::map m_kerning; + + // built texture + List m_atlas; + + public: + static const uint32_t* ASCII; + + // Note: + // charset is a list of range pairs, until a 0 terminator (ex. 32,128,0) + + SpriteFont(); + SpriteFont(const char* file, float size); + SpriteFont(const char* file, float size, const uint32_t* charset); + SpriteFont(const Font& font, float size); + SpriteFont(const Font& font, float size, const uint32_t* charset); + SpriteFont(const SpriteFont&) = delete; + SpriteFont(SpriteFont&& src) noexcept; + ~SpriteFont(); + + void dispose(); + + SpriteFont& operator=(const SpriteFont&) = delete; + SpriteFont& operator=(SpriteFont&& src) noexcept; + + const String& name() const { return m_name; } + float size() const { return m_size; } + float ascent() const { return m_ascent; } + float descent() const { return m_descent; } + float line_gap() const { return m_line_gap; } + float height() const { return m_ascent - m_descent; } + float line_height() const { return m_ascent - m_descent + m_line_gap; } + + List& textures() { return m_atlas; } + + float width_of(const String& text) const; + float width_of_line(const String& text, int start = 0) const; + float height_of(const String& text) const; + + void build(const char* file, float size, const uint32_t* charset); + void build(const Font& font, float size, const uint32_t* charset); + + float get_kerning(uint32_t codepoint0, uint32_t codepoint1) const; + void set_kerning(uint32_t codepoint0, uint32_t codepoint1, float kerning); + + Character& get_character(uint32_t codepoint) { return m_characters[codepoint]; } + const Character& get_character(uint32_t codepoint) const; + Character& operator[](uint32_t codepoint) { return m_characters[codepoint]; } + const Character& operator[](uint32_t codepoint) const; + }; +} \ No newline at end of file diff --git a/public/blah/drawing/subtexture.cpp b/public/blah/drawing/subtexture.cpp new file mode 100644 index 0000000..f148e7a --- /dev/null +++ b/public/blah/drawing/subtexture.cpp @@ -0,0 +1,41 @@ +#include + +using namespace Blah; + +Subtexture::Subtexture() +{ + +} + +Subtexture::Subtexture(const TextureRef& texture, Rect source, Rect frame) + : texture(texture), source(source), frame(frame) +{ + update(); +} + +void Subtexture::update() +{ + draw_coords[0].x = -frame.x; + draw_coords[0].y = -frame.y; + draw_coords[1].x = -frame.x + source.w; + draw_coords[1].y = -frame.y; + draw_coords[2].x = -frame.x + source.w; + draw_coords[2].y = -frame.y + source.h; + draw_coords[3].x = -frame.x; + draw_coords[3].y = -frame.y + source.h; + + if (texture && texture->is_valid()) + { + float uvx = 1.0f / (float)texture->width(); + float uvy = 1.0f / (float)texture->height(); + + tex_coords[0].x = source.x * uvx; + tex_coords[0].y = source.y * uvy; + tex_coords[1].x = (source.x + source.w ) * uvx; + tex_coords[1].y = source.y * uvy; + tex_coords[2].x = (source.x + source.w) * uvx; + tex_coords[2].y = (source.y + source.h) * uvy; + tex_coords[3].x = source.x * uvx; + tex_coords[3].y = (source.y + source.h) * uvy; + } +} diff --git a/public/blah/drawing/subtexture.h b/public/blah/drawing/subtexture.h new file mode 100644 index 0000000..8d3e8a7 --- /dev/null +++ b/public/blah/drawing/subtexture.h @@ -0,0 +1,24 @@ +#pragma once +#include +#include + +namespace Blah +{ + struct Subtexture + { + TextureRef texture; + Rect source; + Rect frame; + Vec2 draw_coords[4]; // draw coords are automatically assigned through Update method + Vec2 tex_coords[4]; // tex coords are automatically assigned through Update method + + Subtexture(); + Subtexture(const TextureRef& texture, Rect source, Rect frame); + + float width() const { return frame.w; } + float height() const { return frame.h; } + + // updates the draw and tex coords + void update(); + }; +} \ No newline at end of file diff --git a/public/blah/filesystem.cpp b/public/blah/filesystem.cpp new file mode 100644 index 0000000..d686e88 --- /dev/null +++ b/public/blah/filesystem.cpp @@ -0,0 +1,136 @@ +#include +#include + +using namespace Blah; + +bool File::exists(const FilePath& path) +{ + return Internal::Platform::file_exists(path.cstr()); +} + +bool File::remove(const FilePath& path) +{ + return Internal::Platform::file_delete(path.cstr()); +} + +bool Directory::create(const FilePath& path) +{ + return Internal::Platform::dir_create(path.cstr()); +} + +bool Directory::exists(const FilePath& path) +{ + return Internal::Platform::dir_exists(path.cstr()); +} + +bool Directory::remove(const FilePath& path) +{ + return Internal::Platform::dir_delete(path.cstr()); +} + +List Directory::enumerate(const FilePath& path, bool recursive) +{ + List list; + + // get files + Internal::Platform::dir_enumerate(list, path.cstr(), recursive); + + // normalize path names + for (auto& it : list) + it.replace('\\', '/'); + + return list; +} + +void Directory::explore(const FilePath& path) +{ + Internal::Platform::dir_explore(path); +} + +FilePath Path::get_file_name(const FilePath& path) +{ + const char* cstr = path.cstr(); + for (int n = path.length() - 2; n >= 0; n --) + if (*(cstr + n) == '/') + { + return FilePath(cstr + n + 1); + } + + return path; +} + +FilePath Path::get_file_name_no_ext(const FilePath& path) +{ + FilePath filename = get_file_name(path); + return get_path_no_ext(filename); +} + +FilePath Path::get_path_no_ext(const FilePath& path) +{ + const char* cstr = path.cstr(); + for (int n = 0; n < path.length(); n++) + if (*(cstr + n) == '.') + return FilePath(cstr, cstr + n); + return path; +} + +FilePath Path::get_directory_name(const FilePath& path) +{ + FilePath directory = path; + while (directory.ends_with("/")) + directory = directory.substr(0, -1); + int last = directory.last_index_of('/'); + if (last >= 0) + directory = directory.substr(0, last + 1); + return directory; +} + +FilePath Path::get_path_after(const FilePath& path, const FilePath& after) +{ + for (int i = 0; i < path.length(); i++) + { + FilePath substr(path.begin() + i, path.end()); + if (substr.starts_with(after, false)) + return FilePath(path.begin() + i + after.length(), path.end()); + } + + return path; +} + +FilePath Path::normalize(const FilePath& path) +{ + FilePath normalized; + + int len = path.length(); + for (int n = 0; n < len; n++) + { + // normalize slashes + if (path[n] == '\\' || path[n] == '/') + { + if (normalized.length() == 0 || normalized[normalized.length() - 1] != '/') + normalized.append('/'); + } + // move up a directory + else if (path[n] == '.' && n < len - 1 && path[n + 1] == '.') + { + // search backwards for last / + bool couldMoveUp = false; + for (int k = normalized.length() - 2; k >= 0; k--) + if (normalized[k] == '/') + { + normalized = normalized.substr(0, k); + couldMoveUp = true; + break; + } + + if (!couldMoveUp) + normalized.append('.'); + else + n++; + } + else + normalized.append(path[n]); + } + + return normalized; +} \ No newline at end of file diff --git a/public/blah/filesystem.h b/public/blah/filesystem.h new file mode 100644 index 0000000..328a1a9 --- /dev/null +++ b/public/blah/filesystem.h @@ -0,0 +1,41 @@ +#pragma once +#include +#include + +namespace Blah +{ + using FilePath = StrOf<265>; + + enum class FileMode + { + None = 0, + Read = 1 << 1, + Write = 1 << 2, + ReadWrite = Read | Write, + }; + + namespace Directory + { + bool create(const FilePath& path); + bool exists(const FilePath& path); + bool remove(const FilePath& path); + List enumerate(const FilePath& str, bool recursive = true); + void explore(const FilePath& path); + } + + namespace File + { + bool exists(const FilePath& path); + bool remove(const FilePath& path); + } + + namespace Path + { + FilePath get_file_name(const FilePath& path); + FilePath get_file_name_no_ext(const FilePath& path); + FilePath get_path_no_ext(const FilePath& path); + FilePath get_path_after(const FilePath& path, const FilePath& after); + FilePath get_directory_name(const FilePath& path); + FilePath normalize(const FilePath& path); + } +} \ No newline at end of file diff --git a/public/blah/graphics/framebuffer.h b/public/blah/graphics/framebuffer.h new file mode 100644 index 0000000..d5a3785 --- /dev/null +++ b/public/blah/graphics/framebuffer.h @@ -0,0 +1,36 @@ +#pragma once +#include +#include +#include +#include + +namespace Blah +{ + typedef StackList Attachments; + + class FrameBuffer + { + public: + virtual ~FrameBuffer() = default; + + // Gets an Attachment from the FrameBuffer + virtual Attachments& attachments() = 0; + + // Gets an Attachment from the FrameBuffer + virtual const Attachments& attachments() const = 0; + + // Gets the width of the FrameBuffer + virtual int width() const = 0; + + // Gets the height of the FrameBuffer + virtual int height() const = 0; + + // Returns true if the FrameBuffer is valid + virtual bool is_valid() const = 0; + + // Destroys the given FrameBuffer + virtual void dispose() = 0; + }; + + typedef std::shared_ptr FrameBufferRef; +} \ No newline at end of file diff --git a/public/blah/graphics/graphics.cpp b/public/blah/graphics/graphics.cpp new file mode 100644 index 0000000..421bf0b --- /dev/null +++ b/public/blah/graphics/graphics.cpp @@ -0,0 +1,340 @@ +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +using namespace Blah; + +namespace +{ + // active graphics device + Internal::GraphicsDevice* device; + + // active graphics device info + Internal::GraphicsDeviceInfo* device_info; + + // list of possible device info + Internal::GraphicsDeviceInfo* device_options[] = + { + &Internal::OpenGL_DeviceInfo + }; +} + +const BlendMode BlendMode::Normal = BlendMode( + BlendOp::Add, + BlendFactor::One, + BlendFactor::OneMinusSrcAlpha, + BlendOp::Add, + BlendFactor::One, + BlendFactor::OneMinusSrcAlpha, + BlendMask::RGBA, + 0xffffffff +); + +GfxAPI Internal::Graphics::pick_api() +{ + for (int i = 0; i < (int)GfxAPI::Count; i++) + { + if (device_options[i]->supported()) + return device_options[i]->api; + } + + return GfxAPI::Any; +} + +bool Internal::Graphics::init(GfxAPI api) +{ + for (int i = 0; i < (int)GfxAPI::Count; i++) + { + if (device_options[i]->api == api) + { + device_info = device_options[i]; + device = device_info->create(); + if (device != nullptr) + { + device->startup(); + if (device->valid) + break; + + device_info->destroy(device); + device = nullptr; + } + } + } + + return device != nullptr && device->valid; +} + +const GraphicsInfo* Graphics::info() +{ + if (device == nullptr || !device->valid) + return nullptr; + return &device->info; +} + +void Internal::Graphics::shutdown() +{ + if (device != nullptr && device_info != nullptr) + { + device->shutdown(); + device_info->destroy(device); + device = nullptr; + device_info = nullptr; + } +} + +void Internal::Graphics::frame() +{ + if (device != nullptr && device->valid) + device->update(); +} + +void Internal::Graphics::before_render() +{ + if (device != nullptr && device->valid) + device->before_render(); +} + +void Internal::Graphics::after_render() +{ + if (device != nullptr && device->valid) + device->after_render(); +} + +TextureRef Graphics::create_texture(const Image& image) +{ + auto tex = create_texture(image.width, image.height, TextureFormat::RGBA); + tex->set_data((unsigned char*)image.pixels); + return tex; +} + +TextureRef Graphics::create_texture(int width, int height, unsigned char* rgba) +{ + auto tex = create_texture(width, height, TextureFormat::RGBA); + tex->set_data(rgba); + return tex; +} + +TextureRef Graphics::create_texture(int width, int height, TextureFormat format) +{ + BLAH_ASSERT(width > 0 && height > 0, "Texture width and height must be larger than 0"); + BLAH_ASSERT((int)format > (int)TextureFormat::None && (int)format < (int)TextureFormat::Count, "Invalid texture format"); + BLAH_ASSERT(device != nullptr && device->valid, "The graphics device has not been created"); + + return device->create_texture(width, height, TextureFilter::Linear, TextureWrap::Repeat, TextureWrap::Repeat, format); +} + +TextureRef Graphics::create_texture(Stream& stream) +{ + Image img = Image(stream); + + if (img.pixels != nullptr && img.width > 0 && img.height > 0) + { + auto tex = create_texture(img.width, img.height, TextureFormat::RGBA); + tex->set_data((unsigned char*)img.pixels); + return tex; + } + + return TextureRef(); +} + +TextureRef Graphics::create_texture(const char* file) +{ + Image img = Image(file); + + if (img.pixels != nullptr) + { + auto tex = create_texture(img.width, img.height, TextureFormat::RGBA); + tex->set_data((unsigned char*)img.pixels); + return tex; + } + + return TextureRef(); +} + +FrameBufferRef Graphics::create_framebuffer(int width, int height) +{ + static const TextureFormat attachment = TextureFormat::RGBA; + return create_framebuffer(width, height, &attachment, 1); +} + +FrameBufferRef Graphics::create_framebuffer(int width, int height, const TextureFormat* attachments, int attachment_count) +{ + BLAH_ASSERT(width > 0 && height > 0, "FrameBuffer width and height must be larger than 0"); + BLAH_ASSERT(attachment_count <= BLAH_MAX_FRAMEBUFFER_ATTACHMENTS, "Exceeded maximum attachment count"); + BLAH_ASSERT(attachment_count > 0, "At least one attachment must be provided"); + for (int i = 0; i < attachment_count; i++) + BLAH_ASSERT((int)attachments[i] > (int)TextureFormat::None && (int)attachments[i] < (int)TextureFormat::Count, "Invalid texture format"); + BLAH_ASSERT(device != nullptr && device->valid, "The graphics device has not been created"); + + return device->create_framebuffer(width, height, attachments, attachment_count); +} + +ShaderRef Graphics::create_shader(const ShaderData* data) +{ + BLAH_ASSERT(device != nullptr && device->valid, "The graphics device has not been created"); + return device->create_shader(data); +} + +MaterialRef Graphics::create_material(const ShaderRef& shader) +{ + BLAH_ASSERT(device != nullptr && device->valid, "The graphics device has not been created"); + BLAH_ASSERT(shader && shader->is_valid(), "The provided shader is invalid"); + + // TODO: + // use a pool for Materials + return MaterialRef(new Material(shader)); +} + +MeshRef Graphics::create_mesh() +{ + BLAH_ASSERT(device != nullptr && device->valid, "Graphics device has not been created"); + return device->create_mesh(); +} + +RenderCall::RenderCall() +{ + blend = BlendMode::Normal; + target = Graphics::backbuffer; + mesh = MeshRef(); + material = MaterialRef(); + has_viewport = false; + has_scissor = false; + viewport = Rect(); + scissor = Rect(); + index_start = 0; + index_count = 0; + instance_count = 0; + depth = Compare::None; + cull = Cull::None; +} + +void Graphics::render(const RenderCall& render_call) +{ + BLAH_ASSERT(device != null && device->valid, "Graphics device has not been created"); + + // Validate Material + if (!render_call.material || !render_call.material->is_valid()) + { + Log::warn("Trying to draw with an invalid Material"); + return; + } + + // Validate Shader + if (!render_call.material->shader() || !render_call.material->shader()->is_valid()) + { + Log::warn("Trying to draw with an invalid Shader"); + return; + } + + // Validate Mesh + if (!render_call.mesh || !render_call.mesh->is_valid()) + { + Log::warn("Trying to draw with an invalid Mesh"); + return; + } + + // Validate FrameBuffer + if (render_call.target && !render_call.target->is_valid()) + { + Log::warn("Trying to draw with an invalid FrameBuffer"); + return; + } + + // copy call + RenderCall call = render_call; + + // Validate Index Count + int64_t index_count = call.mesh->index_count(); + if (call.index_start + call.index_count > index_count) + { + Log::warn( + "Trying to draw more indices than exist in the index buffer (%i-%i / %i)", + call.index_start, + call.index_start + call.index_count, + index_count); + + if (call.index_start > call.index_count) + return; + + call.index_count = call.index_count - call.index_start; + } + + // Validate Instance Count + int64_t instance_count = call.mesh->instance_count(); + if (call.instance_count > instance_count) + { + Log::warn( + "Trying to draw more instances than exist in the index buffer (%i / %i)", + call.instance_count, + instance_count); + + call.instance_count = instance_count; + } + + // Validate Viewport + if (!call.has_viewport) + { + Point drawSize; + + if (!call.target) + drawSize = Point(App::draw_width(), App::draw_height()); + else + drawSize = Point(call.target->width(), call.target->height()); + + call.viewport.x = 0; + call.viewport.y = 0; + call.viewport.w = (float)drawSize.x; + call.viewport.h = (float)drawSize.y; + } + else + { + if (call.viewport.x < 0) + { + call.viewport.w += call.viewport.x; + call.viewport.x = 0; + } + if (call.viewport.y < 0) + { + call.viewport.h += call.viewport.y; + call.viewport.y = 0; + } + if (call.viewport.w < 0) + call.viewport.w = 0; + if (call.viewport.h < 0) + call.viewport.h = 0; + } + + // Validate Scissor + if (call.has_scissor) + { + if (call.scissor.x < 0) + { + call.scissor.w += call.scissor.x; + call.scissor.x = 0; + } + if (call.scissor.y < 0) + { + call.scissor.h += call.scissor.y; + call.scissor.y = 0; + } + if (call.scissor.w < 0) + call.scissor.w = 0; + if (call.scissor.h < 0) + call.scissor.h = 0; + } + + device->render(&call); +} + +void Graphics::clear(const FrameBufferRef& target, uint32_t rgba) +{ + BLAH_ASSERT(device != null && device->valid, "Graphics device has not been created"); + device->clear(target, rgba); +} \ No newline at end of file diff --git a/public/blah/graphics/graphics.h b/public/blah/graphics/graphics.h new file mode 100644 index 0000000..3555a05 --- /dev/null +++ b/public/blah/graphics/graphics.h @@ -0,0 +1,349 @@ +#pragma once +#include +#include +#include +#include +#include +#include + +#define BLAH_ATTACHMENTS 16 +#define BLAH_UNIFORMS 16 +#define BLAH_ATTRIBUTES 16 +#define BLAH_ATTRIBUTE_NAME 32 +#define BLAH_UNIFORM_NAME 32 + +namespace Blah +{ + class Stream; + struct Image; + + class Texture; + typedef std::shared_ptr TextureRef; + + class FrameBuffer; + typedef std::shared_ptr FrameBufferRef; + + class Material; + typedef std::shared_ptr MaterialRef; + + class Shader; + typedef std::shared_ptr ShaderRef; + + class Mesh; + typedef std::shared_ptr MeshRef; + + struct GraphicsInfo + { + GfxAPI api; + bool instancing; + bool origin_bottom_left; + int max_texture_size; + }; + + enum class TextureFilter + { + None, + Linear, + Nearest + }; + + enum class TextureWrap + { + None, + Clamp, + Repeat + }; + + enum class TextureFormat + { + None, + R, + RG, + RGB, + RGBA, + DepthStencil, + Count + }; + + enum class Compare + { + None, + Always, + Never, + Less, + Equal, + LessOrEqual, + Greater, + NotEqual, + GreatorOrEqual + }; + + enum class Cull + { + None = 0, + Front = 1, + Back = 2, + Both = 3 + }; + + enum class BlendOp + { + Add, + Subtract, + ReverseSubtract, + Min, + Max + }; + + enum class BlendFactor + { + Zero, + One, + SrcColor, + OneMinusSrcColor, + DstColor, + OneMinusDstColor, + SrcAlpha, + OneMinusSrcAlpha, + DstAlpha, + OneMinusDstAlpha, + ConstantColor, + OneMinusConstantColor, + ConstantAlpha, + OneMinusConstantAlpha, + SrcAlphaSaturate, + Src1Color, + OneMinusSrc1Color, + Src1Alpha, + OneMinusSrc1Alpha + }; + + enum class BlendMask + { + None = 0, + Red = 1, + Green = 2, + Blue = 4, + Alpha = 8, + RGB = Red | Green | Blue, + RGBA = Red | Green | Blue | Alpha, + }; + + struct BlendMode + { + BlendOp colorOp; + BlendFactor colorSrc; + BlendFactor colorDst; + BlendOp alphaOp; + BlendFactor alphaSrc; + BlendFactor alphaDst; + BlendMask mask; + uint32_t rgba; + + BlendMode() = default; + + BlendMode(BlendOp op, BlendFactor src, BlendFactor dst) + { + colorOp = op; + colorSrc = src; + colorDst = dst; + alphaOp = op; + alphaSrc = src; + alphaDst = dst; + mask = BlendMask::RGBA; + rgba = 0xffffffff; + } + + BlendMode(BlendOp rgbOp, BlendFactor rgbSrc, BlendFactor rgbDst, BlendOp aOp, BlendFactor aSrc, BlendFactor aDst, BlendMask blendMask, uint32_t blendColor) + { + colorOp = rgbOp; + colorSrc = rgbSrc; + colorDst = rgbDst; + alphaOp = aOp; + alphaSrc = aSrc; + alphaDst = aDst; + mask = blendMask; + rgba = blendColor; + } + + inline bool operator==(const BlendMode& rhs) const { + return colorOp == rhs.colorOp && colorSrc == rhs.colorSrc && colorDst == rhs.colorDst && + alphaOp == rhs.alphaOp && alphaSrc == rhs.alphaSrc && alphaDst == rhs.alphaDst && + mask == rhs.mask && rgba == rhs.rgba; + } + inline bool operator!=(const BlendMode& rhs) const { return !(*this == rhs); } + + static const BlendMode Normal; + }; + + enum class UniformType + { + None, + Float, + Float2, + Float3, + Float4, + Mat3x2, + Mat4x4, + Texture + }; + + enum class VertexSemantics + { + None, + Position, + Normal, + Bitangent, + Color0, + Color1, + Color2, + Color3, + Indices, + Weight, + Texcoord0, + Texcoord1, + Texcoord2, + Texcoord3, + Texcoord4, + Texcoord5, + Texcoord6, + Texcoord7 + }; + + enum class VertexAttributeType + { + None, + Byte, + Short, + Int, + Float + }; + + struct VertexAttribute + { + int index; + VertexSemantics semantics; + VertexAttributeType type; + int components; + bool normalized; + }; + + struct ShaderData + { + const char* vertex; + const char* fragment; + }; + + struct ShaderUniform + { + StrOf name; + UniformType type; + int array_length; + }; + + struct ShaderAttribute + { + StrOf name; + StrOf semantic_name; + int semantic_location; + }; + + struct RenderCall + { + // Framebuffer to draw to + FrameBufferRef target; + + // Mesh to draw with + MeshRef mesh; + + // Material to draw with + MaterialRef material; + + // Whether the Render Call should use a specific viewport + bool has_viewport; + + // Whether the Render Call should use a scissor rectangle + bool has_scissor; + + // The viewport (only used if hasViewport is true) + Rect viewport; + + // The scissor rectangle (only used if hasScissor is true) + Rect scissor; + + // First index in the Mesh to draw from + int64_t index_start; + + // Total amount of indices to draw from the Mesh + int64_t index_count; + + // Total amount of instances to draw from the Mesh + int64_t instance_count; + + // Depth Compare Function + Compare depth; + + // Cull Mode + Cull cull; + + // Blend Mode + BlendMode blend; + + // Initializes a default Render Call + RenderCall(); + }; + + namespace Graphics + { + const FrameBufferRef backbuffer; + + // Gets graphics information from the graphics device + const GraphicsInfo* info(); + + // Creates a new Texture. + // If the Texture creation fails, it will return an invalid TextureRef. + TextureRef create_texture(const Image& image); + + // Creates a new Texture. + // If the Texture creation fails, it will return an invalid TextureRef. + TextureRef create_texture(int width, int height, unsigned char* rgba); + + // Creates a new Texture. + // If the Texture creation fails, it will return an invalid TextureRef. + TextureRef create_texture(int width, int height, TextureFormat format); + + // Creates a new Texture from a Stream. + // If the Texture creation fails, it will return an invalid TextureRef. + TextureRef create_texture(Stream& stream); + + // Creates a new Texture from a File. + // If the Texture creation fails, it will return an invalid TextureRef. + TextureRef create_texture(const char* file); + + // Creates a new FrameBuffer with a single Color attachment + // If the FrameBuffer creation fails, it will return an invalid FrameBufferRef. + FrameBufferRef create_framebuffer(int width, int height); + + // Creates a new FrameBuffer with the given Texture Attachments. You must provide at least one Attachment. + // If the FrameBuffer creation fails, it will return an invalid FrameBufferRef. + FrameBufferRef create_framebuffer(int width, int height, const TextureFormat* attachments, int attachmentCount); + + // Creates a Shader with the given Shader Data. + // If the Shader creation fails, it will return an invalid ShaderRef. + ShaderRef create_shader(const ShaderData* data); + + // Creates a new Material from the given Shader. + // If the Shader is invalid, it will return an invalid MaterialRef. + MaterialRef create_material(const ShaderRef& shader); + + // Creates a new Mesh. + // If the Mesh creation fails, it will return an invalid Mesh. + MeshRef create_mesh(); + + // Submits and Flushes the given render call to the graphics device. + void render(const RenderCall& render_call); + + // Clears the given FrameBuffer to the given color. + void clear(const FrameBufferRef& target, uint32_t rgba); + } +} \ No newline at end of file diff --git a/public/blah/graphics/material.cpp b/public/blah/graphics/material.cpp new file mode 100644 index 0000000..eebbc8a --- /dev/null +++ b/public/blah/graphics/material.cpp @@ -0,0 +1,252 @@ +#include + +using namespace Blah; + +namespace +{ + size_t calc_uniform_size(const ShaderUniform& uniform) + { + size_t components = 0; + + switch (uniform.type) + { + case UniformType::Float: components = 1; break; + case UniformType::Float2: components = 2; break; + case UniformType::Float3: components = 3; break; + case UniformType::Float4: components = 4; break; + case UniformType::Mat3x2: components = 6; break; + case UniformType::Mat4x4: components = 16; break; + default: + BLAH_ERROR("Unespected Uniform Type"); + break; + } + + return components * uniform.array_length; + } +} + +Material::Material(const ShaderRef& shader) +{ + m_shader = shader; + m_data = nullptr; + m_disposed = false; + + // invalid shader + if (!m_shader || !m_shader->is_valid()) + { + m_disposed = true; + return; + } + + Uniforms uniforms = shader->uniforms(); + StackList float_offsets; + size_t float_size = 0; + + for (auto& uniform : uniforms) + { + if (uniform.type == UniformType::None) + continue; + + if (uniform.type == UniformType::Texture) + { + for (int i = 0; i < uniform.array_length; i ++) + m_textures.add(TextureRef()); + continue; + } + + float_offsets.add(float_size); + float_size += calc_uniform_size(uniform); + } + + m_data = new float[float_size]; + memset(m_data, 0, sizeof(float) * float_size); + + for (auto& it : float_offsets) + m_floats.add(m_data + it); +} + +Material::~Material() +{ + dispose(); +} + +const ShaderRef Material::shader() const +{ + return m_shader; +} + +void Material::set_texture(const char* name, const TextureRef& texture, int index) +{ + BLAH_ASSERT(!m_disposed, "Material has been disposed"); + BLAH_ASSERT(m_shader && m_shader->is_valid(), "Material Shader is invalid"); + + if (m_textures.count() > 0) + { + int offset = 0; + for (auto& uniform : m_shader->uniforms()) + { + if (uniform.type != UniformType::Texture) + continue; + + if (strcmp(uniform.name, name) == 0) + { + m_textures[offset + index] = texture; + return; + } + + offset += uniform.array_length; + if (offset + index >= m_textures.count()) + break; + } + } + + Log::warn("No Texture Uniform '%s' at index [%i] exists", name, index); +} + +TextureRef Material::get_texture(const char* name, int index) const +{ + BLAH_ASSERT(!m_disposed, "Material has been disposed"); + BLAH_ASSERT(m_shader && m_shader->is_valid(), "Material Shader is invalid"); + + int offset = 0; + for (auto& uniform : m_shader->uniforms()) + { + if (uniform.type != UniformType::Texture) + continue; + + if (strcmp(uniform.name, name) == 0) + return m_textures[offset + index]; + + offset += uniform.array_length; + if (offset + index >= m_textures.count()) + break; + } + + Log::warn("No Texture Uniform '%s' at index [%i] exists", name, index); + return TextureRef(); +} + +TextureRef Material::get_texture(int slot, int index) const +{ + BLAH_ASSERT(!m_disposed, "Material has been disposed"); + BLAH_ASSERT(m_shader && m_shader->is_valid(), "Material Shader is invalid"); + + int offset = 0; + int s = 0; + for (auto& uniform : m_shader->uniforms()) + { + if (uniform.type == UniformType::Texture) + { + if (s == slot) + return m_textures[offset + index]; + + offset += uniform.array_length; + if (offset + index >= m_textures.count()) + break; + } + + s++; + } + + Log::warn("No Texture Uniform ['%i'] at index [%i] exists", slot, index); + return TextureRef(); +} + +void Material::set_value(const char* name, const float* value, int64_t length) +{ + BLAH_ASSERT(!m_disposed, "Material has been disposed"); + BLAH_ASSERT(m_shader && m_shader->is_valid(), "Material Shader is invalid"); + BLAH_ASSERT(length >= 0, "Length must be >= 0"); + + int index = 0; + for (auto& uniform : m_shader->uniforms()) + { + if (uniform.type == UniformType::Texture || uniform.type == UniformType::None) + continue; + + if (strcmp(uniform.name, name) == 0) + { + size_t max = calc_uniform_size(uniform); + if (length > max) + { + Log::warn("Exceeding length of Uniform '%s' (%i / %i)", name, length, max); + length = max; + } + + memcpy(m_floats[index], value, sizeof(float) * length); + return; + } + + index++; + } + + Log::warn("No Uniform '%s' exists", name); +} + +const float* Material::get_value(const char* name, int64_t* length) const +{ + BLAH_ASSERT(!m_disposed, "Material has been disposed"); + BLAH_ASSERT(m_shader && m_shader->is_valid(), "Material Shader is invalid"); + + int index = 0; + for (auto& uniform : m_shader->uniforms()) + { + if (uniform.type == UniformType::Texture || uniform.type == UniformType::None) + continue; + + if (strcmp(uniform.name, name) == 0) + { + if (length != nullptr) + *length = calc_uniform_size(uniform); + return m_floats[index]; + } + + index++; + } + + *length = 0; + return nullptr; + Log::warn("No Uniform '%s' exists", name); +} + +const float* Material::get_value(int slot, int64_t* length) const +{ + BLAH_ASSERT(!m_disposed, "Material has been disposed"); + BLAH_ASSERT(m_shader && m_shader->is_valid(), "Material Shader is invalid"); + + int index = 0; + int s = 0; + for (auto& uniform : m_shader->uniforms()) + { + if (uniform.type != UniformType::Texture && uniform.type != UniformType::None) + { + if (index == slot) + { + if (length != nullptr) + *length = calc_uniform_size(uniform); + return m_floats[index]; + } + index++; + } + + s++; + } + + *length = 0; + return nullptr; + Log::warn("No Uniform [%i] exists", slot); +} + +bool Material::is_valid() const +{ + return !m_disposed && m_shader && m_shader->is_valid(); +} + +void Material::dispose() +{ + delete[] m_data; + m_data = nullptr; + m_shader.reset(); + m_textures.clear(); + m_floats.clear(); +} \ No newline at end of file diff --git a/public/blah/graphics/material.h b/public/blah/graphics/material.h new file mode 100644 index 0000000..80289ff --- /dev/null +++ b/public/blah/graphics/material.h @@ -0,0 +1,66 @@ +#pragma once +#include +#include +#include +#include + +namespace Blah +{ + class Material + { + public: + Material(const ShaderRef& shader); + ~Material(); + + Material(const Material& src) = delete; + Material(Material&& src) = delete; + Material& operator=(const Material& src) = delete; + Material& operator=(Material&& src) = delete; + + // Returns the Shader assigned to the Material. + const ShaderRef shader() const; + + // Sets the texture. If the index is out of bounds, the Material is not valid, the Uniform + // is not a Texture Uniform, or the Uniform does not exist, this will do nothing. + void set_texture(const char* name, const TextureRef& texture, int index = 0); + + // Gets the texture. + // If the Uniform does not exist, or the index is out of bounds, this will return + // an invalid Texture reference. + TextureRef get_texture(const char* name, int index = 0) const; + + // Gets the texture from the given slot. + // If the slot is not a Texture Uniform, or the index is out of bounds, this will return + // an invalid Texture reference. + TextureRef get_texture(int slot, int index = 0) const; + + // Sets the value. Length is the total amount of values to set. For example, if the Uniform type + // is a float2, and there are 4 elements, the maximum length should be 8. + void set_value(const char* name, const float* value, int64_t length); + + // Gets a pointer to the values of the given Uniform, or nullptr if it doesn't exist. + // Length is the total amount of values. For example, if the Uniform type + // is a float2, and there are 4 elements, the length should be 8. + const float* get_value(const char* name, int64_t* length = nullptr) const; + + // Gets a pointer to the values of the given Uniform, or nullptr if it doesn't exist. + // Length is the total amount of values. For example, if the Uniform type + // is a float2, and there are 4 elements, the length should be 8. + const float* get_value(int slot, int64_t* length = nullptr) const; + + // Returns true if the Material is valid + bool is_valid() const; + + // Destroys the Material + void dispose(); + + private: + ShaderRef m_shader; + List m_textures; + List m_floats; + float* m_data; + bool m_disposed; + }; + + typedef std::shared_ptr MaterialRef; +} \ No newline at end of file diff --git a/public/blah/graphics/mesh.h b/public/blah/graphics/mesh.h new file mode 100644 index 0000000..9c76fb9 --- /dev/null +++ b/public/blah/graphics/mesh.h @@ -0,0 +1,47 @@ +#pragma once +#include +#include +#include + +namespace Blah +{ + class Mesh + { + public: + virtual ~Mesh() = default; + + // Sets the Vertex Format of the Mesh + virtual void vertex_format(const VertexAttribute* attributes, int attribute_count, int stride = -1) = 0; + + // Sets the Instance Format of the Mesh + virtual void instance_format(const VertexAttribute* attributes, int attribute_count, int stride = -1) = 0; + + // Uploads the given index buffer to the Mesh + virtual void index_data(const void* indices, int64_t count) = 0; + + // Uploads the given vertex buffer to the Mesh + // Note you must call vertex_format at least once before uploading vertices. + virtual void vertex_data(const void* vertices, int64_t count) = 0; + + // Uploads the given instance buffer to the Mesh + // Note you must call instance_format at least once before uploading instances. + virtual void instance_data(const void* instances, int64_t count) = 0; + + // Gets the index count of the Mesh + virtual int64_t index_count() const = 0; + + // Gets the vertex count of the Mesh + virtual int64_t vertex_count() const = 0; + + // Gets the instance count of the Mesh + virtual int64_t instance_count() const = 0; + + // Returns true if the Mesh is valid + virtual bool is_valid() const = 0; + + // Destroys the given Mesh + virtual void dispose() = 0; + }; + + typedef std::shared_ptr MeshRef; +} \ No newline at end of file diff --git a/public/blah/graphics/shader.h b/public/blah/graphics/shader.h new file mode 100644 index 0000000..2d185cc --- /dev/null +++ b/public/blah/graphics/shader.h @@ -0,0 +1,36 @@ +#pragma once +#include +#include +#include + +namespace Blah +{ + typedef StackList Uniforms; + typedef StackList Attributes; + + class Shader + { + public: + virtual ~Shader() = default; + + // Gets a list of Shader Uniforms from Shader + virtual Uniforms& uniforms() = 0; + + // Gets a list of Shader Uniforms from Shader + virtual const Uniforms& uniforms() const = 0; + + // Gets a list of Shader Attributes from Shader + virtual Attributes& attributes() = 0; + + // Gets a list of Shader Attributes from Shader + virtual const Attributes& attributes() const = 0; + + // Returns true if the Shader is valid + virtual bool is_valid() const = 0; + + // Destroys the given Shader + virtual void dispose() = 0; + }; + + typedef std::shared_ptr ShaderRef; +} \ No newline at end of file diff --git a/public/blah/graphics/texture.h b/public/blah/graphics/texture.h new file mode 100644 index 0000000..197b127 --- /dev/null +++ b/public/blah/graphics/texture.h @@ -0,0 +1,57 @@ +#pragma once +#include +#include + +namespace Blah +{ + class Texture + { + public: + virtual ~Texture() = default; + + // gets the width of the texture + virtual int width() const = 0; + + // gets the height of the texture + virtual int height() const = 0; + + // Gets the format of the Texture + virtual TextureFormat format() const = 0; + + // Sets the filter of the Texture + virtual void set_filter(TextureFilter filter) = 0; + + // Gets the filter of the Texture + virtual TextureFilter get_filter() const = 0; + + // Sets the wrap of the Texture + virtual void set_wrap(TextureWrap x, TextureWrap y) = 0; + + // Gets the wrap of the Texture + virtual TextureWrap get_wrap_x() const = 0; + + // Gets the wrap of the Texture + virtual TextureWrap get_wrap_y() const = 0; + + // Sets the data of the Texture. + // Note that the pixel buffer should be in the same format as the Texture. There is no row padding. + // If the pixel buffer isn't the same size as the texture, it will set the minimum available amount of data. + virtual void set_data(unsigned char* data) = 0; + + // Gets the data of the Texture. + // Note that the pixel buffer will be written to in the same format as the Texture, + // and you should allocate enough space for the full texture. There is no row padding. + virtual void get_data(unsigned char* data) = 0; + + // Returns true if the Texture is part of a FrameBuffer + virtual bool is_framebuffer() const = 0; + + // Returns true if the Texture + virtual bool is_valid() const = 0; + + // Destroys the given Texture + virtual void dispose() = 0; + }; + + typedef std::shared_ptr TextureRef; +} \ No newline at end of file diff --git a/public/blah/images/aseprite.cpp b/public/blah/images/aseprite.cpp new file mode 100644 index 0000000..20c5b6d --- /dev/null +++ b/public/blah/images/aseprite.cpp @@ -0,0 +1,466 @@ +#include +#include + +#define STBI_NO_STDIO +#define STBI_NO_JPEG +#define STBI_NO_PNG +#define STBI_NO_BMP +#define STBI_NO_PSD +#define STBI_NO_TGA +#define STBI_NO_GIF +#define STBI_NO_HDR +#define STBI_NO_PIC +#define STBI_NO_PNM +#include + +#define MAX(a, b) ((a) > (b) ? (a) : (b)) +#define MIN(a, b) ((a) < (b) ? (a) : (b)) +#define MUL_UN8(a, b, t) \ + ((t) = (a) * (uint16_t)(b) + 0x80, ((((t) >> 8) + (t) ) >> 8)) + +using namespace Blah; + +Aseprite::Aseprite() +{ + +} + +Aseprite::Aseprite(const char* path) +{ + FileStream fs(path, FileMode::Read); + Parse(fs); +} + +Aseprite::Aseprite(Stream& stream) +{ + Parse(stream); +} + +Aseprite::Aseprite(const Aseprite& src) +{ + mode = src.mode; + width = src.width; + height = src.height; + layers = src.layers; + frames = src.frames; + tags = src.tags; + slices = src.slices; + palette = src.palette; +} + +Aseprite& Aseprite::operator=(const Aseprite& src) +{ + mode = src.mode; + width = src.width; + height = src.height; + layers = src.layers; + frames = src.frames; + tags = src.tags; + slices = src.slices; + palette = src.palette; + return *this; +} + +Aseprite::Aseprite(Aseprite&& src) noexcept +{ + mode = src.mode; + width = src.width; + height = src.height; + layers = std::move(src.layers); + frames = std::move(src.frames); + tags = std::move(src.tags); + slices = std::move(src.slices); + palette = std::move(src.palette); +} + +Aseprite& Aseprite::operator=(Aseprite&& src) noexcept +{ + mode = src.mode; + width = src.width; + height = src.height; + layers = std::move(src.layers); + frames = std::move(src.frames); + tags = std::move(src.tags); + slices = std::move(src.slices); + palette = std::move(src.palette); + return *this; +} + +Aseprite::~Aseprite() +{ + +} + +void Aseprite::Parse(Stream& stream) +{ + if (!stream.is_readable()) + { + BLAH_ERROR("Stream is not readable"); + return; + } + + int frameCount = 0; + + // header + { + // filesize + stream.read(Endian::Little); + + // magic number + auto magic = stream.read(Endian::Little); + if (magic != 0xA5E0) + { + BLAH_ERROR("File is not a valid Aseprite file"); + return; + } + + // main info + frameCount = stream.read(Endian::Little); + width = stream.read(Endian::Little); + height = stream.read(Endian::Little); + mode = static_cast(stream.read(Endian::Little) / 8); + + // don't care about other info + stream.read(Endian::Little); // Flags + stream.read(Endian::Little); // Speed (deprecated) + stream.read(Endian::Little); // Should be 0 + stream.read(Endian::Little); // Should be 0 + stream.read(Endian::Little); // Palette entry + stream.seek(stream.position() + 3); // Ignore these bytes + stream.read(Endian::Little); // Number of colors (0 means 256 for old sprites) + stream.read(Endian::Little); // Pixel width + stream.read(Endian::Little); // Pixel height + stream.seek(stream.position() + 92); // For Future + } + + frames.expand(frameCount); + + // frames + for (int i = 0; i < frameCount; i ++) + { + auto frameStart = stream.position(); + auto frameEnd = frameStart + stream.read(Endian::Little); + unsigned int chunks = 0; + + // frame header + { + auto magic = stream.read(Endian::Little); // magic number + if (magic != 0xF1FA) + { + BLAH_ERROR("File is not a valid Aseprite file"); + return; + } + + auto old_chunk_count = stream.read(Endian::Little); + frames[i].duration = stream.read(Endian::Little); + stream.seek(stream.position() + 2); // for future + auto new_chunk_count = stream.read(Endian::Little); + + if (old_chunk_count == 0xFFFF) + chunks = new_chunk_count; + else + chunks = old_chunk_count; + } + + // make frame image + frames[i].image = Image(width, height); + + // frame chunks + for (unsigned int j = 0; j < chunks; j ++) + { + auto chunkStart = stream.position(); + auto chunkEnd = chunkStart + stream.read(Endian::Little); + auto chunkType = static_cast(stream.read(Endian::Little)); + + switch (chunkType) + { + case Chunks::Layer: ParseLayer(stream, i); break; + case Chunks::Cel: ParseCel(stream, i, (size_t)chunkEnd); break; + case Chunks::Palette: ParsePalette(stream, i); break; + case Chunks::UserData: ParseUserData(stream, i); break; + case Chunks::FrameTags: ParseTag(stream, i); break; + case Chunks::Slice: ParseSlice(stream, i); break; + default: break; + } + + stream.seek(chunkEnd); + } + + stream.seek(frameEnd); + } +} + +void Aseprite::ParseLayer(Stream& stream, int frame) +{ + auto layer = layers.expand(1); + layer->flag = static_cast(stream.read(Endian::Little)); + layer->visible = ((int)layer->flag & (int)LayerFlags::Visible) == (int)LayerFlags::Visible; + layer->type = static_cast(stream.read(Endian::Little)); + layer->child_level = stream.read(Endian::Little); + stream.read(Endian::Little); // width + stream.read(Endian::Little); // height + layer->blendmode = stream.read(Endian::Little); + layer->alpha = stream.read(Endian::Little); + stream.seek(stream.position() + 3); // for future + + layer->name.set_length(stream.read(Endian::Little)); + stream.read(layer->name.cstr(), layer->name.length()); + + layer->userdata.color = 0xffffff; + layer->userdata.text = ""; + last_userdata = &(layer->userdata); +} + +void Aseprite::ParseCel(Stream& stream, int frameIndex, size_t maxPosition) +{ + Frame& frame = frames[frameIndex]; + + auto cel = frame.cels.expand(1); + cel->layer_index = stream.read(Endian::Little); + cel->x = stream.read(Endian::Little); + cel->y = stream.read(Endian::Little); + cel->alpha = stream.read(Endian::Little); + cel->linked_frame_index = -1; + cel->linked_cel_index = -1; + + auto celType = stream.read(Endian::Little); + stream.seek(stream.position() + 7); + + // RAW or DEFLATE + if (celType == 0 || celType == 2) + { + auto width = stream.read(Endian::Little); + auto height = stream.read(Endian::Little); + auto count = width * height * (int)mode; + + cel->image = Image(width, height); + + // RAW + if (celType == 0) + { + stream.read(cel->image.pixels, count); + } + // DEFLATE (zlib) + else + { + // this could be optimized to use a buffer on the stack if we only read set chunks at a time + // stbi's zlib doesn't have that functionality though + auto size = maxPosition - stream.position(); + if (size > INT32_MAX) + size = INT32_MAX; + if (size < 0) + size = 0; + + char* buffer = new char[size]; + stream.read(buffer, size); + + int olen = width * height * sizeof(Color); + int res = stbi_zlib_decode_buffer((char*)cel->image.pixels, olen, buffer, (int)size); + + delete[] buffer; + + if (res < 0) + { + BLAH_ERROR("Unable to parse Aseprite file"); + return; + } + } + + // convert to pixels + // note: we work in-place to save having to store stuff in a buffer + if (mode == Modes::Grayscale) + { + auto src = (unsigned char*)cel->image.pixels; + auto dst = cel->image.pixels; + for (int d = width * height - 1, s = (width * height - 1) * 2; d >= 0; d --, s -= 2) + dst[d] = Color(src[s], src[s], src[s], src[s + 1]); + } + else if (mode == Modes::Indexed) + { + auto src = (unsigned char*)cel->image.pixels; + auto dst = cel->image.pixels; + for (int i = width * height - 1; i >= 0; i --) + dst[i] = palette[src[i]]; + } + + } + // REFERENCE + // this cel directly references a previous cel + else if (celType == 1) + { + cel->linked_frame_index = stream.read(Endian::Little); + cel->linked_cel_index = static_cast(frame.cels.count() - 1); + } + + // draw to frame if visible + if ((int)layers[cel->layer_index].flag & (int)LayerFlags::Visible) + { + RenderCel(cel, &frame); + } + + cel->userdata.color = 0xffffff; + cel->userdata.text = ""; + last_userdata = &(cel->userdata); +} + +void Aseprite::ParsePalette(Stream& stream, int frame) +{ + /* size */ stream.read(Endian::Little); + auto start = stream.read(Endian::Little); + auto end = stream.read(Endian::Little); + stream.seek(stream.position() + 8); + + palette.expand(palette.count() + (end - start + 1)); + + for (int p = 0, len = static_cast(end - start) + 1; p < len; p++) + { + auto hasName = stream.read(Endian::Little); + + palette[start + p] = stream.read(Endian::Little); + + if (hasName & 0xF000) + { + int len = stream.read(Endian::Little); + stream.seek(stream.position() + len); + } + } +} + +void Aseprite::ParseUserData(Stream& stream, int frame) +{ + if (last_userdata != nullptr) + { + auto flags = stream.read(Endian::Little); + + // has text + if (flags & (1 << 0)) + { + last_userdata->text.set_length(stream.read(Endian::Little)); + stream.read(last_userdata->text.cstr(), last_userdata->text.length()); + } + + // has color + if (flags & (1 << 1)) + last_userdata->color = stream.read(Endian::Little); + } +} + +void Aseprite::ParseTag(Stream& stream, int frame) +{ + auto count = stream.read(Endian::Little); + stream.seek(stream.position() + 8); + + for (int t = 0; t < count; t++) + { + auto tag = tags.expand(1); + tag->from = stream.read(Endian::Little); + tag->to = stream.read(Endian::Little); + tag->loops = static_cast(stream.read(Endian::Little)); + + stream.seek(stream.position() + 8); + tag->color = Color(stream.read(), stream.read(), stream.read(Endian::Little), 255); + stream.seek(stream.position() + 1); + + tag->name.set_length(stream.read(Endian::Little)); + stream.read(tag->name .cstr(), tag->name.length()); + } +} + +void Aseprite::ParseSlice(Stream& stream, int frame) +{ + int count = stream.read(Endian::Little); + int flags = stream.read(Endian::Little); + stream.read(Endian::Little); // reserved + + String name; + name.set_length(stream.read(Endian::Little)); + stream.read(name.cstr(), name.length()); + + for (int s = 0; s < count; s++) + { + auto slice = slices.expand(1); + slice->name = name; + slice->frame = stream.read(Endian::Little); + slice->origin.x = stream.read(Endian::Little); + slice->origin.y = stream.read(Endian::Little); + slice->width = stream.read(Endian::Little); + slice->height = stream.read(Endian::Little); + + // 9 slice (ignored atm) + if (flags & (1 << 0)) + { + stream.read(Endian::Little); + stream.read(Endian::Little); + stream.read(Endian::Little); + stream.read(Endian::Little); + } + + // pivot point + slice->has_pivot = false; + if (flags & (1 << 1)) + { + slice->has_pivot = true; + slice->pivot.x = stream.read(Endian::Little); + slice->pivot.y = stream.read(Endian::Little); + } + + slice->userdata.color = 0xffffff; + slice->userdata.text = ""; + last_userdata = &(slice->userdata); + } +} + +void Aseprite::RenderCel(Cel* cel, Frame* frame) +{ + Layer& layer = layers[cel->layer_index]; + + while (cel->linked_frame_index >= 0) + cel = &(frames[cel->linked_frame_index].cels[cel->linked_cel_index]); + + int t; + unsigned char opacity = MUL_UN8(cel->alpha, layer.alpha, t); + if (opacity <= 0) + return; + + auto src = cel->image.pixels; + auto srcX = cel->x; + auto srcY = cel->y; + auto srcW = cel->image.width; + auto srcH = cel->image.height; + auto dst = frame->image.pixels; + auto dstW = frame->image.width; + auto dstH = frame->image.height; + + // blit pixels + int left = MAX(0, srcX); + int right = MIN(dstW, srcX + srcW); + int top = MAX(0, srcY); + int bottom = MIN(dstH, srcY + srcH); + + if (layer.blendmode == 0) + { + for (int dx = left, sx = -MIN(srcX, 0); dx < right; dx++, sx++) + { + for (int dy = top, sy = -MIN(srcY, 0); dy < bottom; dy++, sy++) + { + Color* srcColor = (src + sx + sy * srcW); + Color* dstColor = (dst + dx + dy * dstW); + + if (srcColor->a != 0) + { + auto sa = MUL_UN8(srcColor->a, opacity, t); + auto ra = dstColor->a + sa - MUL_UN8(dstColor->a, sa, t); + + dstColor->r = (unsigned char)(dstColor->r + (srcColor->r - dstColor->r) * sa / ra); + dstColor->g = (unsigned char)(dstColor->g + (srcColor->g - dstColor->g) * sa / ra); + dstColor->b = (unsigned char)(dstColor->b + (srcColor->b - dstColor->b) * sa / ra); + dstColor->a = (unsigned char)ra; + } + } + } + } + else + { + BLAH_ERROR("Aseprite blendmodes aren't implemented"); + } +} diff --git a/public/blah/images/aseprite.h b/public/blah/images/aseprite.h new file mode 100644 index 0000000..5b76f5a --- /dev/null +++ b/public/blah/images/aseprite.h @@ -0,0 +1,152 @@ +#pragma once +#include +#include +#include +#include +#include + +namespace Blah +{ + class Aseprite + { + public: + + enum class Modes + { + Indexed = 1, + Grayscale = 2, + RGBA = 4 + }; + + enum class Chunks + { + OldPaletteA = 0x0004, + OldPaletteB = 0x0011, + Layer = 0x2004, + Cel = 0x2005, + CelExtra = 0x2006, + Mask = 0x2016, + Path = 0x2017, + FrameTags = 0x2018, + Palette = 0x2019, + UserData = 0x2020, + Slice = 0x2022 + }; + + enum class LoopDirections + { + Forward = 0, + Reverse = 1, + PingPong = 2 + }; + + enum class LayerFlags + { + Visible = 1, + Editable = 2, + LockMovement = 4, + Background = 8, + PreferLinkedCels = 16, + Collapsed = 32, + Reference = 64 + }; + + enum class LayerTypes + { + Normal = 0, + Group = 1 + }; + + struct UserData + { + String text; + Color color; + }; + + struct Layer; + + struct Cel + { + int layer_index = 0; + int linked_frame_index = 0; + int linked_cel_index = 0; + int x = 0; + int y = 0; + unsigned char alpha = 0; + Image image; + UserData userdata; + }; + + struct Frame + { + int duration = 0; + Image image; + List cels; + }; + + struct Layer + { + LayerFlags flag = (LayerFlags)0; + LayerTypes type = LayerTypes::Normal; + String name; + int child_level = 0; + int blendmode = 0; + unsigned char alpha = 0; + bool visible = true; + UserData userdata; + }; + + struct Tag + { + String name; + LoopDirections loops = LoopDirections::Forward; + int from = 0; + int to = 0; + Color color; + UserData userdata; + }; + + struct Slice + { + int frame = 0; + String name; + Point origin; + int width = 0; + int height = 0; + bool has_pivot = false; + Point pivot; + UserData userdata; + }; + + Modes mode = Modes::RGBA; + int width = 0; + int height = 0; + + List layers; + List frames; + List tags; + List slices; + List palette; + + Aseprite(); + Aseprite(const char* path); + Aseprite(Stream& stream); + Aseprite(const Aseprite& src); + Aseprite& operator=(const Aseprite& src); + Aseprite(Aseprite&& src) noexcept; + Aseprite& operator=(Aseprite&& src) noexcept; + ~Aseprite(); + + private: + UserData* last_userdata = nullptr; + + void Parse(Stream& stream); + void ParseLayer(Stream& stream, int frame); + void ParseCel(Stream& stream, int frame, size_t maxPosition); + void ParsePalette(Stream& stream, int frame); + void ParseUserData(Stream& stream, int frame); + void ParseTag(Stream& stream, int frame); + void ParseSlice(Stream& stream, int frame); + void RenderCel(Cel* cel, Frame* frame); + }; +} \ No newline at end of file diff --git a/public/blah/images/font.cpp b/public/blah/images/font.cpp new file mode 100644 index 0000000..a947d81 --- /dev/null +++ b/public/blah/images/font.cpp @@ -0,0 +1,244 @@ +#include +#include +#include +#include + +using namespace Blah; + +#define STBTT_STATIC +#define STB_TRUETYPE_IMPLEMENTATION +#include + +String GetName(stbtt_fontinfo* font, int nameId) +{ + int length = 0; + + // get the name + const uint16_t* ptr = (const uint16_t*)stbtt_GetFontNameStr(font, &length, + STBTT_PLATFORM_ID_MICROSOFT, + STBTT_MS_EID_UNICODE_BMP, + STBTT_MS_LANG_ENGLISH, + nameId); + + // we want the size in wide chars + length /= 2; + + String str; + if (length > 0) + str.append_utf16(ptr, ptr + length, Calc::is_little_endian()); + return str; +} + +Font::Font() +{ + font = nullptr; + data = nullptr; + ascent = 0; + descent = 0; + lineGap = 0; + valid = false; +} + +Font::Font(Stream& stream) : Font() +{ + Load(stream); +} + +Font::Font(const char* path) : Font() +{ + FileStream fs(path, FileMode::Read); + if (fs.is_readable()) + Load(fs); +} + +Font::Font(Font&& src) noexcept +{ + font = src.font; + data = src.data; + familyName = src.familyName; + styleName = src.styleName; + ascent = src.ascent; + descent = src.descent; + lineGap = src.lineGap; + valid = src.valid; + + src.familyName.clear(); + src.styleName.clear(); + src.valid = false; + src.font = nullptr; + src.data = nullptr; +} + +Font& Font::operator=(Font&& src) noexcept +{ + font = src.font; + data = src.data; + familyName = src.familyName; + styleName = src.styleName; + ascent = src.ascent; + descent = src.descent; + lineGap = src.lineGap; + valid = src.valid; + + src.familyName.clear(); + src.styleName.clear(); + src.valid = false; + src.font = nullptr; + src.data = nullptr; + return *this; +} + +Font::~Font() +{ + dispose(); +} + +void Font::Load(Stream& stream) +{ + dispose(); + + if (!stream.is_readable()) + { + BLAH_ERROR("Unable to load a font as the Stream was not readable"); + return; + } + + // create data buffer + auto size = stream.length(); + data = new unsigned char[size]; + stream.read(data, size); + + // init font + font = new stbtt_fontinfo(); + auto fn = (stbtt_fontinfo*)font; + stbtt_InitFont(fn, data, 0); + familyName = GetName(fn, 1); + styleName = GetName(fn, 2); + + // properties + stbtt_GetFontVMetrics(fn, &ascent, &descent, &lineGap); + valid = true; +} + +void Font::dispose() +{ + delete (stbtt_fontinfo*)font; + delete[] data; + font = nullptr; + data = nullptr; + familyName.dispose(); + styleName.dispose(); +} + +const char* Font::FamilyName() const +{ + return familyName.cstr(); +} + +const char* Font::StyleName() const +{ + return styleName.cstr(); +} + +int Font::Ascent() const +{ + return ascent; +} + +int Font::Descent() const +{ + return descent; +} + +int Font::LineGap() const +{ + return lineGap; +} + +int Font::Height() const +{ + return ascent - descent; +} + +int Font::LineHeight() const +{ + return ascent - descent + lineGap; +} + +int Font::GetGlyph(Codepoint codepoint) const +{ + if (font == nullptr) + return 0; + return stbtt_FindGlyphIndex((stbtt_fontinfo*)font, codepoint); +} + +float Font::GetScale(float size) const +{ + if (font == nullptr) + return 0; + return stbtt_ScaleForPixelHeight((stbtt_fontinfo*)font, size); +} + +float Font::GetKerning(int glyph1, int glyph2, float scale) const +{ + if (font == nullptr) + return 0; + return stbtt_GetGlyphKernAdvance((stbtt_fontinfo*)font, glyph1, glyph2) * scale; +} + +Font::Char Font::GetCharacter(int glyph, float scale) const +{ + Char ch; + + if (font == nullptr) + return ch; + + int advance, offsetX, x0, y0, x1, y1; + + stbtt_GetGlyphHMetrics((stbtt_fontinfo*)font, glyph, &advance, &offsetX); + stbtt_GetGlyphBitmapBox((stbtt_fontinfo*)font, glyph, scale, scale, &x0, &y0, &x1, &y1); + + int w = (x1 - x0); + int h = (y1 - y0); + + // define character + ch.glyph = glyph; + ch.width = w; + ch.height = h; + ch.advance = advance * scale; + ch.offsetX = offsetX * scale; + ch.offsetY = (float)y0; + ch.scale = scale; + ch.hasGlyph = (w > 0 && h > 0 && stbtt_IsGlyphEmpty((stbtt_fontinfo*)font, glyph) == 0); + + return ch; +} + +bool Font::GetBitmap(const Font::Char& ch, Color* pixels) const +{ + if (ch.hasGlyph) + { + // we actually use the image buffer as our temporary buffer, and fill the pixels out backwards after + // kinda weird but it works & saves creating more memory + unsigned char* src = (unsigned char*)pixels; + stbtt_MakeGlyphBitmap((stbtt_fontinfo*)font, src, ch.width, ch.height, ch.width, ch.scale, ch.scale, ch.glyph); + + int len = ch.width * ch.height; + for (int a = (len - 1) * 4, b = (len - 1); b >= 0; a -= 4, b -= 1) + { + src[a + 0] = src[b]; + src[a + 1] = src[b]; + src[a + 2] = src[b]; + src[a + 3] = src[b]; + } + + return true; + } + + return false; +} + +bool Font::IsValid() const +{ + return valid; +} \ No newline at end of file diff --git a/public/blah/images/font.h b/public/blah/images/font.h new file mode 100644 index 0000000..5d7352c --- /dev/null +++ b/public/blah/images/font.h @@ -0,0 +1,62 @@ +#pragma once +#include +#include +#include + +namespace Blah +{ + typedef uint32_t Codepoint; + + class Font + { + public: + struct Char + { + int glyph = 0; + int width = 0; + int height = 0; + float advance = 0; + float offsetX = 0; + float offsetY = 0; + float scale = 0; + bool hasGlyph = false; + }; + + Font(); + Font(Stream& stream); + Font(const char* path); + Font(const Font&) = delete; + Font& operator=(const Font&) = delete; + Font(Font&& src) noexcept; + Font& operator=(Font&& src) noexcept; + ~Font(); + + void dispose(); + + const char* FamilyName() const; + const char* StyleName() const; + int Ascent() const; + int Descent() const; + int LineGap() const; + int Height() const; + int LineHeight() const; + + int GetGlyph(Codepoint codepoint) const; + float GetScale(float size) const; + float GetKerning(int glyph1, int glyph2, float scale) const; + Char GetCharacter(int glyph, float scale) const; + bool GetBitmap(const Char& ch, Color* pixels) const; + bool IsValid() const; + + private: + void Load(Stream& stream); + void* font; + unsigned char* data; + String familyName; + String styleName; + int ascent; + int descent; + int lineGap; + bool valid; + }; +} \ No newline at end of file diff --git a/public/blah/images/image.cpp b/public/blah/images/image.cpp new file mode 100644 index 0000000..6fef463 --- /dev/null +++ b/public/blah/images/image.cpp @@ -0,0 +1,284 @@ +#include +#include +#include +#include + +using namespace Blah; + +#define STB_IMAGE_IMPLEMENTATION +#define STBI_ONLY_JPEG +#define STBI_ONLY_PNG +#define STBI_ONLY_BMP +#include + +#define STB_IMAGE_WRITE_IMPLEMENTATION +#include + +int Engine_STBI_Read(void *user, char *data, int size) +{ + int64_t read = ((Stream*)user)->read(data, size); + return (int)read; +} + +void Engine_STBI_Skip(void *user, int n) +{ + ((Stream*)user)->seek(((Stream*)user)->position() + n); +} + +int Engine_STBI_Eof(void *user) +{ + int64_t position = ((Stream*)user)->position(); + int64_t length = ((Stream*)user)->length(); + + if (position >= length) + return 1; + + return 0; +} + +void Engine_STBI_Write(void *context, void *data, int size) +{ + ((Stream*)context)->write((char*)data, size); +} + +Image::Image() +{ + width = height = 0; + pixels = nullptr; +} + +Image::Image(Stream& stream) +{ + width = height = 0; + pixels = nullptr; + from_stream(stream); +} + +Image::Image(const char* file) +{ + width = height = 0; + pixels = nullptr; + + FileStream fs(file, FileMode::Read); + if (fs.is_readable()) + from_stream(fs); +} + +Image::Image(int width, int height) +{ + BLAH_ASSERT(width >= 0 && height >= 0, "Image width and height must be larger than 0"); + + this->width = width; + this->height = height; + pixels = new Color[width * height]; + memset(pixels, 0, (size_t)width * (size_t)height * sizeof(Color)); +} + +Image::Image(const Image& src) +{ + width = src.width; + height = src.height; + pixels = nullptr; + + if (src.pixels != nullptr && width > 0 && height > 0) + { + pixels = new Color[width * height]; + memcpy(pixels, src.pixels, sizeof(Color) * width * height); + } +} + +Image& Image::operator=(const Image& src) +{ + width = src.width; + height = src.height; + pixels = nullptr; + + if (src.pixels != nullptr && width > 0 && height > 0) + { + pixels = new Color[width * height]; + memcpy(pixels, src.pixels, sizeof(Color) * width * height); + } + + return *this; +} + +Image::Image(Image&& src) noexcept +{ + width = src.width; + height = src.height; + pixels = src.pixels; + src.width = src.height = 0; + src.pixels = nullptr; +} + +Image& Image::operator=(Image&& src) noexcept +{ + width = src.width; + height = src.height; + pixels = src.pixels; + src.width = src.height = 0; + src.pixels = nullptr; + return *this; +} + +Image::~Image() +{ + dispose(); +} + +void Image::from_stream(Stream& stream) +{ + dispose(); + + if (!stream.is_readable()) + { + BLAH_ERROR("Unable to load image as the Stream was not readable"); + return; + } + + stbi_io_callbacks callbacks; + callbacks.eof = Engine_STBI_Eof; + callbacks.read = Engine_STBI_Read; + callbacks.skip = Engine_STBI_Skip; + + int x, y, comps; + uint8_t* data = stbi_load_from_callbacks(&callbacks, &stream, &x, &y, &comps, 4); + + if (data == nullptr) + { + BLAH_ERROR("Unable to load image as the Stream's data was not a valid image"); + return; + } + + width = x; + height = y; + pixels = (Color*)data; + +} + +void Image::dispose() +{ + delete[] pixels; + pixels = nullptr; + width = height = 0; +} + +void Image::premultiply() +{ + if (pixels != nullptr) + { + for (int n = 0; n < width * height; n ++) + { + pixels[n].r = (uint8_t)(pixels[n].r * pixels[n].a / 255); + pixels[n].g = (uint8_t)(pixels[n].g * pixels[n].a / 255); + pixels[n].b = (uint8_t)(pixels[n].b * pixels[n].a / 255); + } + } +} + +void Image::set_pixels(const RectI& rect, Color* data) +{ + for (int y = 0; y < rect.h; y++) + { + int to = rect.x + ((rect.y + y) * width); + int from = (y * rect.w); + memcpy(pixels + to, data + from, sizeof(Color) * rect.w); + } +} + +bool Image::save_png(const char* file) const +{ + FileStream fs(file, FileMode::Write); + return save_png(fs); +} + +bool Image::save_png(Stream& stream) const +{ + BLAH_ASSERT(pixels != nullptr, "Image Pixel data cannot be null"); + BLAH_ASSERT(width > 0 && height > 0, "Image Width and Height must be larger than 0"); + + if (stream.is_writable()) + { + stbi_write_force_png_filter = 0; + stbi_write_png_compression_level = 0; + + if (stbi_write_png_to_func(Engine_STBI_Write, &stream, width, height, 4, pixels, width * 4) != 0) + return true; + else + Log::error("stbi_write_png_to_func failed"); + } + else + { + Log::error("Cannot save Image, the Stream is not writable"); + } + + return false; +} + +bool Image::save_jpg(const char* file, int quality) const +{ + FileStream fs(file, FileMode::Write); + return save_jpg(fs, quality); +} + +bool Image::save_jpg(Stream& stream, int quality) const +{ + BLAH_ASSERT(pixels != nullptr, "Image Pixel data cannot be null"); + BLAH_ASSERT(width > 0 && height > 0, "Image Width and Height must be larger than 0"); + + if (quality < 1) + { + Log::warn("jpg quality value should be between 1 and 100; input was %i", quality); + quality = 1; + } + else if (quality > 100) + { + Log::warn("jpg quality value should be between 1 and 100; input was %i", quality); + quality = 100; + } + + if (stream.is_writable()) + { + if (stbi_write_jpg_to_func(Engine_STBI_Write, &stream, width, height, 4, pixels, quality) != 0) + return true; + else + Log::error("stbi_write_jpg_to_func failed"); + } + else + { + Log::error("Cannot save Image, the Stream is not writable"); + } + + return false; +} + + + +void Image::get_pixels(Color* dest, const Point& destPos, const Point& destSize, RectI sourceRect) +{ + // can't be outside of the source image + if (sourceRect.x < 0) sourceRect.x = 0; + if (sourceRect.y < 0) sourceRect.y = 0; + if (sourceRect.x + sourceRect.w > width) sourceRect.w = width - sourceRect.x; + if (sourceRect.y + sourceRect.h > height) sourceRect.h = height - sourceRect.y; + + // can't be larger than our destination + if (sourceRect.w > destSize.x - destPos.x) + sourceRect.w = destSize.x - destPos.x; + if (sourceRect.h > destSize.y - destPos.y) + sourceRect.h = destSize.y - destPos.y; + + for (int y = 0; y < sourceRect.h; y++) + { + int to = destPos.x + (destPos.y + y) * destSize.x; + int from = sourceRect.x + (sourceRect.y + y) * width; + memcpy(dest + to, pixels + from, sizeof(Color) * (int)sourceRect.w); + } +} + +Image Image::get_sub_image(const RectI& sourceRect) +{ + Image img(sourceRect.w, sourceRect.h); + get_pixels(img.pixels, Point::zero, Point(img.width, img.height), sourceRect); + return img; +} diff --git a/public/blah/images/image.h b/public/blah/images/image.h new file mode 100644 index 0000000..e00e1d8 --- /dev/null +++ b/public/blah/images/image.h @@ -0,0 +1,38 @@ +#pragma once +#include +#include +#include + +namespace Blah +{ + class Stream; + + struct Image + { + int width = 0; + int height = 0; + Color* pixels = nullptr; + + Image(); + Image(Stream& stream); + Image(const char* file); + Image(int width, int height); + Image(const Image& src); + Image& operator=(const Image& src); + Image(Image&& src) noexcept; + Image& operator=(Image&& src) noexcept; + ~Image(); + + void from_stream(Stream& stream); + void dispose(); + + void premultiply(); + void set_pixels(const RectI& rect, Color* data); + bool save_png(const char* file) const; + bool save_png(Stream& stream) const; + bool save_jpg(const char* file, int quality) const; + bool save_jpg(Stream& stream, int quality) const; + void get_pixels(Color* dest, const Point& destPos, const Point& destSize, RectI sourceRect); + Image get_sub_image(const RectI& sourceRect); + }; +} \ No newline at end of file diff --git a/public/blah/images/packer.cpp b/public/blah/images/packer.cpp new file mode 100644 index 0000000..96ce747 --- /dev/null +++ b/public/blah/images/packer.cpp @@ -0,0 +1,336 @@ +#include +#include +#include + +using namespace Blah; + +Packer::Packer() + : max_size(8192), power_of_two(true), spacing(1), padding(1), dirty(false) { } + +Packer::Packer(int max_size, int spacing, bool power_of_two) + : max_size(max_size), power_of_two(power_of_two), spacing(spacing), padding(1), dirty(false) { } + +Packer::Packer(Packer&& src) noexcept +{ + max_size = src.max_size; + power_of_two = src.power_of_two; + spacing = src.spacing; + padding = src.padding; + dirty = src.dirty; + pages = std::move(src.pages); + entries = std::move(src.entries); + buffer = std::move(src.buffer); +} + +Packer& Packer::operator=(Packer&& src) noexcept +{ + max_size = src.max_size; + power_of_two = src.power_of_two; + spacing = src.spacing; + padding = src.padding; + dirty = src.dirty; + pages = std::move(src.pages); + entries = std::move(src.entries); + buffer = std::move(src.buffer); + return *this; +} + +Packer::~Packer() +{ + dispose(); +} + +void Packer::add(uint64_t id, int width, int height, const Color* pixels) +{ + add_entry(id, width, height, pixels); +} + +void Packer::add(uint64_t id, const Image& image) +{ + add_entry(id, image.width, image.height, image.pixels); +} + +void Packer::add(uint64_t id, const String& path) +{ + add(id, Image(path.cstr())); +} + +void Packer::add_entry(uint64_t id, int w, int h, const Color* pixels) +{ + dirty = true; + + Entry* entry = entries.expand(1); + entry->id = id; + entry->page = 0; + entry->empty = true; + entry->frame = RectI(0, 0, w, h); + entry->packed = RectI(0, 0, 0, 0); + + // trim + int top = 0, left = 0, right = w, bottom = h; + + // TOP: + for (int y = 0; y < h; y++) + for (int x = 0, s = y * w; x < w; x++, s++) + if (pixels[s].a > 0) + { + top = y; + goto JUMP_LEFT; + } + JUMP_LEFT: + for (int x = 0; x < w; x++) + for (int y = top, s = x + y * w; y < h; y++, s += w) + if (pixels[s].a > 0) + { + left = x; + goto JUMP_RIGHT; + } + JUMP_RIGHT: + for (int x = w - 1; x >= left; x--) + for (int y = top, s = x + y * w; y < h; y++, s += w) + if (pixels[s].a > 0) + { + right = x + 1; + goto JUMP_BOTTOM; + } + JUMP_BOTTOM: + for (int y = h - 1; y >= top; y--) + for (int x = left, s = x + y * w; x < right; x++, s++) + if (pixels[s].a > 0) + { + bottom = y + 1; + goto JUMP_END; + } + JUMP_END:; + + // pixels actually exist in this source + if (right >= left && bottom >= top) + { + entry->empty = false; + + // store size + entry->frame.x = -left; + entry->frame.y = -top; + entry->packed.w = (right - left); + entry->packed.h = (bottom - top); + + // create pixel data + entry->memoryIndex = buffer.position(); + + // copy pixels over + if (entry->packed.w == w && entry->packed.h == h) + { + buffer.write((char*)pixels, sizeof(Color) * w * h); + } + else + { + for (int i = 0; i < entry->packed.h; i++) + buffer.write((char*)(pixels + left + (top + i) * entry->frame.w), sizeof(Color) * entry->packed.w); + } + } +} + +void Packer::pack() +{ + if (!dirty) + return; + + dirty = false; + pages.clear(); + + // only if we have stuff to pack + auto count = entries.count(); + if (count > 0) + { + // get all the sources sorted largest -> smallest + List sources; + { + sources.expand(count); + int index = 0; + + for (int i = 0; i < entries.count(); i++) + sources[index++] = &entries[i]; + + std::sort(sources.begin(), sources.end(), [](Packer::Entry* a, Packer::Entry* b) + { + return a->packed.w * a->packed.h > b->packed.w * b->packed.h; + }); + } + + // make sure the largest isn't too large + if (sources[0]->packed.w + padding * 2 > max_size || sources[0]->packed.h + padding * 2 > max_size) + { + BLAH_ERROR("Source image is larger than max atlas size"); + return; + } + + // we should never need more nodes than source images * 3 + // if this causes problems we could change it to use push_back I suppose + List nodes; + nodes.expand(count * 4); + + int packed = 0, page = 0; + while (packed < count) + { + if (sources[packed]->empty) + { + packed++; + continue; + } + + int from = packed; + int index = 0; + Node* root = nodes[index++].Reset(RectI(0, 0, sources[from]->packed.w + padding * 2 + spacing, sources[from]->packed.h + padding * 2 + spacing)); + + while (packed < count) + { + if (sources[packed]->empty) + { + packed++; + continue; + } + + int w = sources[packed]->packed.w + padding * 2 + spacing; + int h = sources[packed]->packed.h + padding * 2 + spacing; + + Node* node = root->Find(w, h); + + // try to expand + if (node == nullptr) + { + bool canGrowDown = (w <= root->rect.w) && (root->rect.h + h < max_size); + bool canGrowRight = (h <= root->rect.h) && (root->rect.w + w < max_size); + bool shouldGrowRight = canGrowRight && (root->rect.h >= (root->rect.w + w)); + bool shouldGrowDown = canGrowDown && (root->rect.w >= (root->rect.h + h)); + + if (canGrowDown || canGrowRight) + { + // grow right + if (shouldGrowRight || (!shouldGrowDown && canGrowRight)) + { + Node* next = nodes[index++].Reset(RectI(0, 0, root->rect.w + w, root->rect.h)); + next->used = true; + next->down = root; + next->right = node = nodes[index++].Reset(RectI(root->rect.w, 0, w, root->rect.h)); + root = next; + } + // grow down + else + { + Node* next = nodes[index++].Reset(RectI(0, 0, root->rect.w, root->rect.h + h)); + next->used = true; + next->down = node = nodes[index++].Reset(RectI(0, root->rect.h, root->rect.w, h)); + next->right = root; + root = next; + } + } + } + + // doesn't fit + if (node == nullptr) + break; + + // add + node->used = true; + node->down = nodes[index++].Reset(RectI(node->rect.x, node->rect.y + h, node->rect.w, node->rect.h - h)); + node->right = nodes[index++].Reset(RectI(node->rect.x + w, node->rect.y, node->rect.w - w, h)); + + sources[packed]->packed.x = node->rect.x + padding; + sources[packed]->packed.y = node->rect.y + padding; + packed++; + } + + // get page size + int pageWidth, pageHeight; + if (power_of_two) + { + pageWidth = 2; + pageHeight = 2; + while (pageWidth < root->rect.w) + pageWidth *= 2; + while (pageHeight < root->rect.h) + pageHeight *= 2; + } + else + { + pageWidth = root->rect.w; + pageHeight = root->rect.h; + } + + // create each page + { + pages.add(Image(pageWidth, pageHeight)); + + // copy image data to image + for (int i = from; i < packed; i++) + { + sources[i]->page = page; + if (!sources[i]->empty) + { + RectI dst = sources[i]->packed; + Color* src = (Color*)(buffer.data() + sources[i]->memoryIndex); + + // TODO: + // Optimize this? + if (padding > 0) + { + pages[page].set_pixels(RectI(dst.x - padding, dst.y, dst.w, dst.h), src); + pages[page].set_pixels(RectI(dst.x + padding, dst.y, dst.w, dst.h), src); + pages[page].set_pixels(RectI(dst.x, dst.y - padding, dst.w, dst.h), src); + pages[page].set_pixels(RectI(dst.x, dst.y + padding, dst.w, dst.h), src); + } + + pages[page].set_pixels(dst, src); + + } + } + } + + page++; + } + } +} + +void Packer::clear() +{ + pages.clear(); + entries.clear(); + dirty = false; +} + +void Packer::dispose() +{ + pages.dispose(); + entries.dispose(); + buffer.close(); + max_size = 0; + power_of_two = 0; + spacing = 0; + dirty = false; +} + +Packer::Node::Node() + : used(false), rect(0, 0, 0, 0), right(nullptr), down(nullptr) { } + +Packer::Node* Packer::Node::Find(int w, int h) +{ + if (used) + { + Packer::Node* r = right->Find(w, h); + if (r != nullptr) + return r; + return down->Find(w, h); + } + else if (w <= rect.w && h <= rect.h) + return this; + return nullptr; +} + +Packer::Node* Packer::Node::Reset(const RectI& rect) +{ + used = false; + this->rect = rect; + right = nullptr; + down = nullptr; + return this; +} \ No newline at end of file diff --git a/public/blah/images/packer.h b/public/blah/images/packer.h new file mode 100644 index 0000000..a3cf0f7 --- /dev/null +++ b/public/blah/images/packer.h @@ -0,0 +1,71 @@ +#pragma once +#include +#include +#include +#include +#include +#include +#include + +namespace Blah +{ + + class Packer + { + public: + class Entry + { + friend class Packer; + private: + int64_t memoryIndex = 0; + public: + int page = 0; + uint64_t id; + bool empty = true; + RectI packed; + RectI frame; + }; + + int max_size; + bool power_of_two; + int spacing; + int padding; + + List pages; + List entries; + + Packer(); + Packer(int max_size, int spacing, bool power_of_two); + Packer(const Packer&) = delete; + Packer& operator=(const Packer&) = delete; + Packer(Packer&& src) noexcept; + Packer& operator=(Packer&& src) noexcept; + ~Packer(); + + void add(uint64_t id, int width, int height, const Color* pixels); + void add(uint64_t id, const Image& bitmap); + void add(uint64_t id, const String& path); + + void pack(); + void clear(); + void dispose(); + + private: + struct Node + { + bool used; + RectI rect; + Node* right; + Node* down; + + Node(); + Node* Find(int w, int h); + Node* Reset(const RectI& rect); + }; + + bool dirty; + BufferStream buffer; + + void add_entry(uint64_t id, int w, int h, const Color* pixels); + }; +} \ No newline at end of file diff --git a/public/blah/input/input.cpp b/public/blah/input/input.cpp new file mode 100644 index 0000000..03f5e08 --- /dev/null +++ b/public/blah/input/input.cpp @@ -0,0 +1,378 @@ +#include +#include +#include +#include +#include +#include +#include + +using namespace Blah; + +namespace +{ + InputState g_last_state; + InputState g_curr_state; + InputState g_next_state; + InputState g_empty_state; + ControllerState g_empty_controller; +} + +void Internal::Input::init() +{ + g_empty_controller.name = "Disconnected"; + for (int i = 0; i < BLAH_MAX_CONTROLLERS; i++) + g_empty_state.controllers[i].name = g_empty_controller.name; + + g_last_state = g_empty_state; + g_curr_state = g_empty_state; + g_next_state = g_empty_state; +} + +void Internal::Input::frame() +{ + // cycle states + g_last_state = g_curr_state; + g_curr_state = g_next_state; + + // copy state, clear pressed / released values + { + for (int i = 0; i < BLAH_MAX_KEYBOARD_KEYS; i++) + { + g_next_state.keyboard.pressed[i] = false; + g_next_state.keyboard.released[i] = false; + } + + for (int i = 0; i < BLAH_MAX_MOUSE_BUTTONS; i++) + { + g_next_state.mouse.pressed[i] = false; + g_next_state.mouse.released[i] = false; + } + g_next_state.mouse.wheel = Point::zero; + + for (int i = 0; i < BLAH_MAX_TEXT; i ++) + g_next_state.keyboard.text[i] = 0; + + for (int i = 0; i < BLAH_MAX_CONTROLLERS; i++) + { + ControllerState* controller = &(g_next_state.controllers[i]); + + if (!controller->is_connected) + controller->name = nullptr; + + for (int j = 0; j < BLAH_MAX_CONTROLLER_BUTTONS; j++) + { + controller->pressed[j] = false; + controller->released[j] = false; + } + } + } +} + +void Internal::Input::on_mouse_move(float x, float y) +{ + g_next_state.mouse.position.x = x; + g_next_state.mouse.position.y = y; + + Point size = Point(App::width(), App::height()); + Point draw = Point(App::draw_width(), App::draw_height()); + + g_next_state.mouse.draw_position.x = (x / (float)size.x) * draw.x; + g_next_state.mouse.draw_position.y = (y / (float)size.y) * draw.y; +} + +void Internal::Input::on_mouse_screen_move(float x, float y) +{ + g_next_state.mouse.screen_position.x = x; + g_next_state.mouse.screen_position.y = y; +} + +void Internal::Input::on_mouse_down(MouseButton button) +{ + int i = (int)button; + if (i >= 0 && i < BLAH_MAX_MOUSE_BUTTONS) + { + g_next_state.mouse.down[i] = true; + g_next_state.mouse.pressed[i] = true; + g_next_state.mouse.timestamp[i] = Time::milliseconds; + } +} + +void Internal::Input::on_mouse_up(MouseButton button) +{ + int i = (int)button; + if (i >= 0 && i < BLAH_MAX_MOUSE_BUTTONS) + { + g_next_state.mouse.down[i] = false; + g_next_state.mouse.released[i] = true; + } +} + +void Internal::Input::on_key_down(Key key) +{ + int i = (int)key; + if (i >= 0 && i < BLAH_MAX_KEYBOARD_KEYS) + { + g_next_state.keyboard.down[i] = true; + g_next_state.keyboard.pressed[i] = true; + g_next_state.keyboard.timestamp[i] = Time::milliseconds; + } +} + +void Internal::Input::on_mouse_wheel(Point wheel) +{ + g_next_state.mouse.wheel = wheel; +} + +void Internal::Input::on_key_up(Key key) +{ + int i = (int)key; + if (i >= 0 && i < BLAH_MAX_KEYBOARD_KEYS) + { + g_next_state.keyboard.down[i] = false; + g_next_state.keyboard.released[i] = true; + } +} + +void Internal::Input::on_text_utf8(const char* text) +{ + strncat(g_next_state.keyboard.text, text, BLAH_MAX_TEXT); +} + +void Internal::Input::on_controller_connect(int index, const char* name, int is_gamepad, int button_count, int axis_count) +{ + if (index < BLAH_MAX_CONTROLLERS) + { + ControllerState* controller = &(g_next_state.controllers[index]); + *controller = g_empty_controller; + controller->name = name; + controller->is_connected = 1; + controller->is_gamepad = is_gamepad; + controller->button_count = button_count; + controller->axis_count = axis_count; + } +} + +void Internal::Input::on_controller_disconnect(int index) +{ + if (index < BLAH_MAX_CONTROLLERS) + g_next_state.controllers[index] = g_empty_controller; +} + +void Internal::Input::on_button_down(int index, int button) +{ + if (index < BLAH_MAX_CONTROLLERS && + button < BLAH_MAX_CONTROLLER_BUTTONS && + g_next_state.controllers[index].is_connected && + button < g_next_state.controllers[index].button_count) + { + g_next_state.controllers[index].down[button] = 1; + g_next_state.controllers[index].pressed[button] = 1; + g_next_state.controllers[index].button_timestamp[button] = Time::milliseconds; + } +} + +void Internal::Input::on_button_up(int index, int button) +{ + if (index < BLAH_MAX_CONTROLLERS && + button < BLAH_MAX_CONTROLLER_BUTTONS && + g_next_state.controllers[index].is_connected && + button < g_next_state.controllers[index].button_count) + { + g_next_state.controllers[index].down[button] = 0; + g_next_state.controllers[index].released[button] = 1; + } +} + +void Internal::Input::on_axis_move(int index, int axis, float value) +{ + if (index < BLAH_MAX_CONTROLLERS && + axis < BLAH_MAX_CONTROLLER_AXIS && + g_next_state.controllers[index].is_connected && + axis < g_next_state.controllers[index].axis_count) + { + g_next_state.controllers[index].axis[axis] = value; + g_next_state.controllers[index].axis_timestamp[axis] = Time::milliseconds; + } +} + +const InputState* Input::state() +{ + return &g_curr_state; +} + +const InputState* Input::last_state() +{ + return &g_last_state; +} + + +Vec2 Input::mouse_draw() +{ + return Vec2(g_curr_state.mouse.draw_position); +} + +bool Input::pressed(MouseButton button) +{ + int i = (int)button; + return i >= 0 && i < BLAH_MAX_MOUSE_BUTTONS && g_curr_state.mouse.pressed[i]; +} + +bool Input::down(MouseButton button) +{ + int i = (int)button; + return i >= 0 && i < BLAH_MAX_MOUSE_BUTTONS && g_curr_state.mouse.down[i]; +} + +bool Input::released(MouseButton button) +{ + int i = (int)button; + return i >= 0 && i < BLAH_MAX_MOUSE_BUTTONS && g_curr_state.mouse.released[i]; +} + +Point Input::mouse_wheel() +{ + return g_curr_state.mouse.wheel; +} + +bool Input::pressed(Key key) +{ + int i = (int)key; + return i > 0 && i < BLAH_MAX_KEYBOARD_KEYS && g_curr_state.keyboard.pressed[i]; +} + +bool Input::down(Key key) +{ + int i = (int)key; + return i > 0 && i < BLAH_MAX_KEYBOARD_KEYS && g_curr_state.keyboard.down[i]; +} + +bool Input::released(Key key) +{ + int i = (int)key; + return i > 0 && i < BLAH_MAX_KEYBOARD_KEYS && g_curr_state.keyboard.released[i]; +} + +bool Input::ctrl() +{ + return down(Key::LeftControl) || down(Key::RightControl); +} + +bool Input::shift() +{ + return down(Key::LeftShift) || down(Key::RightShift); +} + +bool Input::alt() +{ + return down(Key::LeftAlt) || down(Key::RightAlt); +} + +const char* Input::text() +{ + return g_curr_state.keyboard.text; +} + +const ControllerState* Input::controller(int controllerIndex) +{ + if (controllerIndex >= BLAH_MAX_CONTROLLERS) + { + Log::warn("Trying to access a controller at %i, outside of EX_MAX_CONTROLLERS", controllerIndex); + return &g_empty_controller; + } + else if (!g_curr_state.controllers[controllerIndex].is_connected) + { + return &g_empty_controller; + } + else + { + return &g_curr_state.controllers[controllerIndex]; + } +} + +bool Input::pressed(int controllerIndex, Button button) +{ + int i = (int)button; + if (controllerIndex < BLAH_MAX_CONTROLLERS && i >= 0 && i < BLAH_MAX_CONTROLLER_BUTTONS) + return g_curr_state.controllers[controllerIndex].pressed[i]; + return false; +} + +bool Input::down(int controllerIndex, Button button) +{ + int i = (int)button; + if (controllerIndex < BLAH_MAX_CONTROLLERS && i >= 0 && i < BLAH_MAX_CONTROLLER_BUTTONS) + return g_curr_state.controllers[controllerIndex].down[i]; + return false; +} + +bool Input::released(int controllerIndex, Button button) +{ + int i = (int)button; + if (controllerIndex < BLAH_MAX_CONTROLLERS && i >= 0 && i < BLAH_MAX_CONTROLLER_BUTTONS) + return g_curr_state.controllers[controllerIndex].released[i]; + return false; +} + +float Input::axis_check(int controllerIndex, Axis axis) +{ + int i = (int)axis; + if (controllerIndex < BLAH_MAX_CONTROLLERS && i >= 0 && i < BLAH_MAX_CONTROLLER_AXIS) + return g_curr_state.controllers[controllerIndex].axis[i]; + return 0; +} + +int Input::axis_check(int prev, Key negative, Key positive) +{ + if (Input::pressed(positive)) + return 1; + else if (Input::pressed(negative)) + return -1; + else + { + bool pos = Input::down(positive); + bool neg = Input::down(negative); + + if (pos && neg) + return prev; + else if (pos) + return 1; + else if (neg) + return -1; + else + return 0; + } +} + +int Input::axis_check(int prev, int controllerIndex, Button negative, Button positive) +{ + if (Input::pressed(controllerIndex, positive)) + return 1; + else if (Input::pressed(controllerIndex, negative)) + return -1; + else + { + bool pos = Input::down(controllerIndex, positive); + bool neg = Input::down(controllerIndex, negative); + + if (pos && neg) + return prev; + else if (pos) + return 1; + else if (neg) + return -1; + else + return 0; + } +} + +const char* Input::name_of(Key key) +{ + switch (key) + { + #define DEFINE_KEY(name, value) case Key::name: return #name; + BLAH_KEY_DEFINITIONS + #undef DEFINE_KEY + } + + return "Unknown"; +} \ No newline at end of file diff --git a/public/blah/input/input.h b/public/blah/input/input.h new file mode 100644 index 0000000..e1de50d --- /dev/null +++ b/public/blah/input/input.h @@ -0,0 +1,431 @@ +#pragma once + +#include +#include + +#define BLAH_MAX_CONTROLLERS 4 +#define BLAH_MAX_CONTROLLER_BUTTONS 64 +#define BLAH_MAX_CONTROLLER_AXIS 16 +#define BLAH_MAX_MOUSE_BUTTONS 16 +#define BLAH_MAX_KEYBOARD_KEYS 512 +#define BLAH_MAX_TEXT 256 +#define BLAH_MAX_VIRTUAL_NODES 32 + +#define BLAH_KEY_DEFINITIONS \ + DEFINE_KEY(Unknown, 0) \ + DEFINE_KEY(A, 4) \ + DEFINE_KEY(B, 5) \ + DEFINE_KEY(C, 6) \ + DEFINE_KEY(D, 7) \ + DEFINE_KEY(E, 8) \ + DEFINE_KEY(F, 9) \ + DEFINE_KEY(G, 10) \ + DEFINE_KEY(H, 11) \ + DEFINE_KEY(I, 12) \ + DEFINE_KEY(J, 13) \ + DEFINE_KEY(K, 14) \ + DEFINE_KEY(L, 15) \ + DEFINE_KEY(M, 16) \ + DEFINE_KEY(N, 17) \ + DEFINE_KEY(O, 18) \ + DEFINE_KEY(P, 19) \ + DEFINE_KEY(Q, 20) \ + DEFINE_KEY(R, 21) \ + DEFINE_KEY(S, 22) \ + DEFINE_KEY(T, 23) \ + DEFINE_KEY(U, 24) \ + DEFINE_KEY(V, 25) \ + DEFINE_KEY(W, 26) \ + DEFINE_KEY(X, 27) \ + DEFINE_KEY(Y, 28) \ + DEFINE_KEY(Z, 29) \ + DEFINE_KEY(D1, 30) \ + DEFINE_KEY(D2, 31) \ + DEFINE_KEY(D3, 32) \ + DEFINE_KEY(D4, 33) \ + DEFINE_KEY(D5, 34) \ + DEFINE_KEY(D6, 35) \ + DEFINE_KEY(D7, 36) \ + DEFINE_KEY(D8, 37) \ + DEFINE_KEY(D9, 38) \ + DEFINE_KEY(D0, 39) \ + DEFINE_KEY(Enter, 40) \ + DEFINE_KEY(Escape, 41) \ + DEFINE_KEY(Backspace, 42) \ + DEFINE_KEY(Tab, 43) \ + DEFINE_KEY(Space, 44) \ + DEFINE_KEY(Minus, 45) \ + DEFINE_KEY(Equals, 46) \ + DEFINE_KEY(LeftBracket, 47) \ + DEFINE_KEY(RightBracket, 48) \ + DEFINE_KEY(BackSlash, 49) \ + DEFINE_KEY(NonUSHash, 50) \ + DEFINE_KEY(Semicolon, 51) \ + DEFINE_KEY(Apostrophe, 52) \ + DEFINE_KEY(Grave, 53) \ + DEFINE_KEY(Comma, 54) \ + DEFINE_KEY(Period, 55) \ + DEFINE_KEY(Slash, 56) \ + DEFINE_KEY(Capslock, 57) \ + DEFINE_KEY(F1, 58) \ + DEFINE_KEY(F2, 59) \ + DEFINE_KEY(F3, 60) \ + DEFINE_KEY(F4, 61) \ + DEFINE_KEY(F5, 62) \ + DEFINE_KEY(F6, 63) \ + DEFINE_KEY(F7, 64) \ + DEFINE_KEY(F8, 65) \ + DEFINE_KEY(F9, 66) \ + DEFINE_KEY(F10, 67) \ + DEFINE_KEY(F11, 68) \ + DEFINE_KEY(F12, 69) \ + DEFINE_KEY(PrintScreen, 70) \ + DEFINE_KEY(ScrollLock, 71) \ + DEFINE_KEY(Pause, 72) \ + DEFINE_KEY(Insert, 73) \ + DEFINE_KEY(Home, 74) \ + DEFINE_KEY(PageUp, 75) \ + DEFINE_KEY(Delete, 76) \ + DEFINE_KEY(End, 77) \ + DEFINE_KEY(PageDown, 78) \ + DEFINE_KEY(Right, 79) \ + DEFINE_KEY(Left, 80) \ + DEFINE_KEY(Down, 81) \ + DEFINE_KEY(Up, 82) \ + DEFINE_KEY(NumlockClear, 83) \ + DEFINE_KEY(KP_Divide, 84) \ + DEFINE_KEY(KP_Multiply, 85) \ + DEFINE_KEY(KP_Minus, 86) \ + DEFINE_KEY(KP_Plus, 87) \ + DEFINE_KEY(KP_Enter, 88) \ + DEFINE_KEY(KP_1, 89) \ + DEFINE_KEY(KP_2, 90) \ + DEFINE_KEY(KP_3, 91) \ + DEFINE_KEY(KP_4, 92) \ + DEFINE_KEY(KP_5, 93) \ + DEFINE_KEY(KP_6, 94) \ + DEFINE_KEY(KP_7, 95) \ + DEFINE_KEY(KP_8, 96) \ + DEFINE_KEY(KP_9, 97) \ + DEFINE_KEY(KP_0, 98) \ + DEFINE_KEY(KP_Period, 99) \ + DEFINE_KEY(NonUSBackSlash, 100) \ + DEFINE_KEY(Application, 101) \ + DEFINE_KEY(Power, 102) \ + DEFINE_KEY(KP_Equals, 103) \ + DEFINE_KEY(F13, 104) \ + DEFINE_KEY(F14, 105) \ + DEFINE_KEY(F15, 106) \ + DEFINE_KEY(F16, 107) \ + DEFINE_KEY(F17, 108) \ + DEFINE_KEY(F18, 109) \ + DEFINE_KEY(F19, 110) \ + DEFINE_KEY(F20, 111) \ + DEFINE_KEY(F21, 112) \ + DEFINE_KEY(F22, 113) \ + DEFINE_KEY(F23, 114) \ + DEFINE_KEY(F24, 115) \ + DEFINE_KEY(Execute, 116) \ + DEFINE_KEY(Help, 117) \ + DEFINE_KEY(Menu, 118) \ + DEFINE_KEY(Select, 119) \ + DEFINE_KEY(Stop, 120) \ + DEFINE_KEY(Again, 121) \ + DEFINE_KEY(Undo, 122) \ + DEFINE_KEY(Cut, 123) \ + DEFINE_KEY(Copy, 124) \ + DEFINE_KEY(Paste, 125) \ + DEFINE_KEY(Find, 126) \ + DEFINE_KEY(Mute, 127) \ + DEFINE_KEY(VolumeUp, 128) \ + DEFINE_KEY(VolumeDown, 129) \ + DEFINE_KEY(KP_Comma, 133) \ + DEFINE_KEY(KP_EqualsAs400, 134) \ + DEFINE_KEY(International1, 135) \ + DEFINE_KEY(International2, 136) \ + DEFINE_KEY(International3, 137) \ + DEFINE_KEY(International4, 138) \ + DEFINE_KEY(International5, 139) \ + DEFINE_KEY(International6, 140) \ + DEFINE_KEY(International7, 141) \ + DEFINE_KEY(International8, 142) \ + DEFINE_KEY(International9, 143) \ + DEFINE_KEY(Language1, 144) \ + DEFINE_KEY(Language2, 145) \ + DEFINE_KEY(Language3, 146) \ + DEFINE_KEY(Language4, 147) \ + DEFINE_KEY(Language5, 148) \ + DEFINE_KEY(Language6, 149) \ + DEFINE_KEY(Language7, 150) \ + DEFINE_KEY(Language8, 151) \ + DEFINE_KEY(Language9, 152) \ + DEFINE_KEY(AltErase, 153) \ + DEFINE_KEY(SysReq, 154) \ + DEFINE_KEY(Cancel, 155) \ + DEFINE_KEY(clear, 156) \ + DEFINE_KEY(Prior, 157) \ + DEFINE_KEY(Return2, 158) \ + DEFINE_KEY(Separator, 159) \ + DEFINE_KEY(Out, 160) \ + DEFINE_KEY(Oper, 161) \ + DEFINE_KEY(ClearAgain, 162) \ + DEFINE_KEY(CRSEL, 163) \ + DEFINE_KEY(EXSEL, 164) \ + DEFINE_KEY(KP_00, 176) \ + DEFINE_KEY(KP_000, 177) \ + DEFINE_KEY(ThousandsSeparator, 178) \ + DEFINE_KEY(DecimalSeparator, 179) \ + DEFINE_KEY(CurrencyUnit, 180) \ + DEFINE_KEY(CurrencySubUnit, 181) \ + DEFINE_KEY(KP_LeftParen, 182) \ + DEFINE_KEY(KP_RightParent, 183) \ + DEFINE_KEY(KP_LeftBrace, 184) \ + DEFINE_KEY(KP_RightBrace, 185) \ + DEFINE_KEY(KP_Tab, 186) \ + DEFINE_KEY(KP_BackSpace, 187) \ + DEFINE_KEY(KP_A, 188) \ + DEFINE_KEY(KP_B, 189) \ + DEFINE_KEY(KP_C, 190) \ + DEFINE_KEY(KP_D, 191) \ + DEFINE_KEY(KP_E, 192) \ + DEFINE_KEY(KP_F, 193) \ + DEFINE_KEY(KP_XOR, 194) \ + DEFINE_KEY(KP_Power, 195) \ + DEFINE_KEY(KP_Percent, 196) \ + DEFINE_KEY(KP_Less, 197) \ + DEFINE_KEY(KP_Greater, 198) \ + DEFINE_KEY(KP_Ampersand, 199) \ + DEFINE_KEY(KP_DoubleAmpersand, 200) \ + DEFINE_KEY(KP_VerticalBar, 201) \ + DEFINE_KEY(KP_DoubleVerticalBar, 202) \ + DEFINE_KEY(KP_Colon, 203) \ + DEFINE_KEY(KP_Hash, 204) \ + DEFINE_KEY(KP_Space, 205) \ + DEFINE_KEY(KP_At, 206) \ + DEFINE_KEY(KP_EXCLAM, 207) \ + DEFINE_KEY(KP_MemStore, 208) \ + DEFINE_KEY(KP_MemRecall, 209) \ + DEFINE_KEY(KP_MemClear, 210) \ + DEFINE_KEY(KP_MemAdd, 211) \ + DEFINE_KEY(KP_MemSubstract, 212) \ + DEFINE_KEY(KP_MemMultiply, 213) \ + DEFINE_KEY(KP_MemDivide, 214) \ + DEFINE_KEY(KP_PlusMinus, 215) \ + DEFINE_KEY(KP_Clear, 216) \ + DEFINE_KEY(KP_ClearEntry, 217) \ + DEFINE_KEY(KP_Binary, 218) \ + DEFINE_KEY(KP_Octal, 219) \ + DEFINE_KEY(KP_Decimal, 220) \ + DEFINE_KEY(KP_Hexadecimal, 221) \ + DEFINE_KEY(LeftControl, 224) \ + DEFINE_KEY(LeftShift, 225) \ + DEFINE_KEY(LeftAlt, 226) \ + DEFINE_KEY(LeftGui, 227) \ + DEFINE_KEY(RightControl, 228) \ + DEFINE_KEY(RightShift, 229) \ + DEFINE_KEY(RightAlt, 230) \ + DEFINE_KEY(RightGui, 231) + +namespace Blah +{ + + struct ControllerState + { + // The name of the gamepad, or NULL if not connected + const char* name; + + // Whether this gamepad is connected + bool is_connected; + + // Whether this gamepad is a standard Game Controller + bool is_gamepad; + + // The total button count for this controller, with a maximum of EX_MAX_CONTROLLER_BUTTONS + int button_count; + + // The total axis count for this controller, with a maximum of EX_MAX_CONTROLLER_AXIS + int axis_count; + + // An array holding the pressed state of each button + bool pressed[BLAH_MAX_CONTROLLER_BUTTONS]; + + // An array holding the down state of each button + bool down[BLAH_MAX_CONTROLLER_BUTTONS]; + + // An array holding the released state of each button + bool released[BLAH_MAX_CONTROLLER_BUTTONS]; + + // An array holding the value state of each axis + float axis[BLAH_MAX_CONTROLLER_AXIS]; + + // Timestamp, in milliseconds, since each button was last pressed + uint64_t button_timestamp[BLAH_MAX_CONTROLLER_BUTTONS]; + + // Timestamp, in milliseconds, since each axis last had a value set + uint64_t axis_timestamp[BLAH_MAX_CONTROLLER_AXIS]; + + }; + + struct KeyboardState + { + bool pressed[BLAH_MAX_KEYBOARD_KEYS]; + bool down[BLAH_MAX_KEYBOARD_KEYS]; + bool released[BLAH_MAX_KEYBOARD_KEYS]; + uint64_t timestamp[BLAH_MAX_KEYBOARD_KEYS]; + char text[BLAH_MAX_TEXT]; + }; + + struct MouseState + { + bool pressed[BLAH_MAX_MOUSE_BUTTONS]; + bool down[BLAH_MAX_MOUSE_BUTTONS]; + bool released[BLAH_MAX_MOUSE_BUTTONS]; + uint64_t timestamp[BLAH_MAX_MOUSE_BUTTONS]; + Vec2 screen_position; + Vec2 draw_position; + Vec2 position; + Point wheel; + }; + + // An Input State, which stores the state for gamepads, keyboard, and mouse + struct InputState + { + // All the Gamepads. Note that not all gamepads are necessarily connected, + // and each one must be checked before use. + ControllerState controllers[BLAH_MAX_CONTROLLERS]; + + // The current Keyboard state + KeyboardState keyboard; + + // The current Mouse state + MouseState mouse; + + }; + + // Keyboard Keys + // These are generally copied from the SDL2 Scancode Keys + enum class Key + { + #define DEFINE_KEY(name, value) name = value, + BLAH_KEY_DEFINITIONS + #undef DEFINE_KEY + }; + + // Game Controller Buttons + enum class Button + { + None = -1, + A = 0, + B = 1, + X = 2, + Y = 3, + Back = 4, + Select = 5, + Start = 6, + LeftStick= 7, + RightStick = 8, + LeftShoulder = 9, + RightShoulder = 10, + Up = 11, + Down = 12, + Left = 13, + Right = 14 + }; + + // Game Controller Axes + enum class Axis + { + None = -1, + LeftX = 0, + LeftY = 1, + RightX = 2, + RightY = 3, + LeftTrigger = 4, + RightTrigger = 5 + }; + + // Mouse Buttons + enum class MouseButton + { + None = -1, + Left = 0, + Middle = 1, + Right = 2, + }; + + namespace Input + { + // Returns the Input State of the current frame. + // This pointer is only valid for the current frame and should not be stored. + const InputState* state(); + + // Returns the Input State of the previous frame. + // This pointer is only valid for the current frame and should not be stored. + const InputState* last_state(); + + // Gets the Mouse Position + Vec2 mouse(); + + // Gets the Draw Mouse Position (Mouse Position / Window Size * Draw Size) + Vec2 mouse_draw(); + + // Gets the Mouse Position in Screen Coordinates + Vec2 mouse_screen(); + + // Checks if the given Mouse Button is pressed + bool pressed(MouseButton button); + + // Checks if the given Mouse Button is down + bool down(MouseButton button); + + // Checks if the given Mouse Button is released + bool released(MouseButton button); + + // Gets the Mouse Wheel + Point mouse_wheel(); + + // Returns 1 if the keyboard key was pressed this frame, and 0 otherwise. + bool pressed(Key key); + + // Returns 1 if the keyboard key was held this frame, and 0 otherwise. + bool down(Key key); + + // Returns 1 if the keyboard key was released this frame, and 0 otherwise. + bool released(Key key); + + // Checks if the Left or Right Ctrl Key is down + bool ctrl(); + + // Checks if the Left or Right Shift Key is down + bool shift(); + + // Checks if the Left or Right Alt Key is down + bool alt(); + + // Get the current Text Input + const char* text(); + + // Gets the controller info for the current controller index. + // If the controller is not connected or the index it out of range, this will set an unconnected controller. + const ControllerState* controller(int controllerIndex); + + // Returns 1 if the button on the controller was pressed this frame, and 0 otherwise. + // If the controller is not connected, or the index is out of range, this will return 0. + bool pressed(int controllerIndex, Button button); + + // Returns 1 if the button on the controller was held this frame, and 0 otherwise. + // If the controller is not connected, or the index is out of range, this will return 0. + bool down(int controllerIndex, Button button); + + // Returns 1 if the button on the controller was released this frame, and 0 otherwise. + // If the controller is not connected, or the index is out of range, this will return 0. + bool released(int controllerIndex, Button button); + + float axis_check(int controllerIndex, Axis axis); + int axis_check(int prev, Key negative, Key positive); + int axis_check(int prev, int controllerIndex, Button negative, Button positive); + + // returns a string name of the given key + const char* name_of(Key key); + } +} \ No newline at end of file diff --git a/public/blah/input/virtual_axis.cpp b/public/blah/input/virtual_axis.cpp new file mode 100644 index 0000000..dc1b888 --- /dev/null +++ b/public/blah/input/virtual_axis.cpp @@ -0,0 +1,4 @@ +#pragma once +#include + +using namespace Blah; \ No newline at end of file diff --git a/public/blah/input/virtual_axis.h b/public/blah/input/virtual_axis.h new file mode 100644 index 0000000..4c88907 --- /dev/null +++ b/public/blah/input/virtual_axis.h @@ -0,0 +1,84 @@ +#pragma once +#include + +namespace Blah +{ + class VirtualAxis + { + private: + struct KeysNode + { + Key positive = Key::Unknown; + Key negative = Key::Unknown; + + int value = 0; + + void init(Key negative, Key positive); + void update(); + }; + + struct ButtonsNode + { + int gamepad_id = 0; + Button positive = Button::None; + Button negative = Button::None; + + int value = 0; + + void init(int gamepad_id, Button negative, Button positive); + void update(); + }; + + struct AxisNode + { + int gamepad_id = 0; + Axis axis = Axis::None; + float deadzone = 0; + + float value = 0; + + void init(int gamepad_id, Axis axis, float deadzone); + void update(); + }; + + KeysNode m_keys[BLAH_MAX_VIRTUAL_NODES]; + ButtonsNode m_buttons[BLAH_MAX_VIRTUAL_NODES]; + AxisNode m_axes[BLAH_MAX_VIRTUAL_NODES]; + int m_keys_len = 0; + int m_buttons_len = 0; + int m_axes_len = 0; + + float m_press_buffer = 0; + float m_release_buffer = 0; + float m_repeat_delay = 0; + float m_repeat_interval = 0; + + float m_value = 0; + int m_value_i = 0; + float m_last_value = 0; + int m_last_value_i = 0; + bool m_pressed = false; + bool m_released = false; + float m_last_press_time = 0; + float m_last_release_time = 0; + float m_repeat_press_time = 0; + + public: + VirtualAxis& add_keys(Key negative, Key positive); + VirtualAxis& add_buttons(int gamepad_id, Button negative, Button positive); + VirtualAxis& add_axis(int gamepad_id, Axis axis, float deadzone); + VirtualAxis& repeat(float m_repeat_delay, float m_repeat_interval); + VirtualAxis& press_buffer(float duration); + VirtualAxis& release_buffer(float duration); + + void update(); + float value() const { return m_value; } + int value_i() const { return m_value_i; } + float last_value() const { return m_last_value; } + int last_value_i() const { return m_last_value_i; } + bool pressed() const { return m_pressed; } + bool released() const { return m_released; } + void clear_press_buffer() { m_last_press_time = -1; m_pressed = false; } + void clear_release_buffer() { m_last_release_time = -1; m_released = false; } + }; +} \ No newline at end of file diff --git a/public/blah/input/virtual_button.cpp b/public/blah/input/virtual_button.cpp new file mode 100644 index 0000000..d290dff --- /dev/null +++ b/public/blah/input/virtual_button.cpp @@ -0,0 +1,4 @@ +#pragma once +#include + +using namespace Blah; \ No newline at end of file diff --git a/public/blah/input/virtual_button.h b/public/blah/input/virtual_button.h new file mode 100644 index 0000000..bf792d1 --- /dev/null +++ b/public/blah/input/virtual_button.h @@ -0,0 +1,83 @@ +#pragma once +#include + +namespace Blah +{ + class VirtualButton + { + private: + struct KeyNode + { + Key key = Key::Unknown; + + bool down = false; + bool pressed = false; + bool released = false; + + void init(Key key); + void update(); + }; + + struct ButtonNode + { + int gamepad_id = 0; + Button button = Button::None; + + bool down = false; + bool pressed = false; + bool released = false; + + void init(int gamepad_id, Button button); + void update(); + }; + + struct AxisNode + { + int gamepad_id = 0; + Axis axis = Axis::None; + float threshold = 0; + bool greater_than = false; + + bool down = false; + bool pressed = false; + bool released = false; + + void init(int gamepad_id, Axis axis, float threshold, bool greater_than); + void update(); + }; + + KeyNode m_keys[BLAH_MAX_VIRTUAL_NODES]; + ButtonNode m_buttons[BLAH_MAX_VIRTUAL_NODES]; + AxisNode m_axes[BLAH_MAX_VIRTUAL_NODES]; + int m_keys_len = 0; + int m_buttons_len = 0; + int m_axes_len = 0; + + float m_press_buffer = 0; + float m_release_buffer = 0; + float m_repeat_delay = 0; + float m_repeat_interval = 0; + + bool m_down = false; + bool m_pressed = false; + bool m_released = false; + float m_last_press_time = 0; + float m_last_release_time = 0; + float m_repeat_press_time = 0; + + public: + VirtualButton& add_key(Key key); + VirtualButton& add_button(int gamepad_id, Button button); + VirtualButton& add_axis(int gamepad_id, Axis axis, float threshold, bool greater_than); + VirtualButton& repeat(float m_repeat_delay, float m_repeat_interval); + VirtualButton& press_buffer(float duration); + VirtualButton& release_buffer(float duration); + + void update(); + bool down() const { return m_down; } + bool pressed() const { return m_pressed; } + bool released() const { return m_released; } + void clear_press_buffer() { m_last_press_time = -1; m_pressed = false; } + void clear_release_buffer() { m_last_release_time = -1; m_released = false; } + }; +} \ No newline at end of file diff --git a/public/blah/input/virtual_stick.cpp b/public/blah/input/virtual_stick.cpp new file mode 100644 index 0000000..4f3490e --- /dev/null +++ b/public/blah/input/virtual_stick.cpp @@ -0,0 +1,4 @@ +#pragma once +#include + +using namespace Blah; \ No newline at end of file diff --git a/public/blah/input/virtual_stick.h b/public/blah/input/virtual_stick.h new file mode 100644 index 0000000..543d89e --- /dev/null +++ b/public/blah/input/virtual_stick.h @@ -0,0 +1,95 @@ +#pragma once +#include +#include +#include + +namespace Blah +{ + class VirtualStick + { + private: + struct KeysNode + { + Key left; + Key right; + Key up; + Key down; + + Point value; + + void init(Key left, Key right, Key up, Key down); + void update(); + }; + + struct ButtonsNode + { + int gamepad_id; + Button left; + Button right; + Button up; + Button down; + + Point value; + + void init(int gamepad_id, Button left, Button right, Button up, Button down); + void update(); + }; + + struct AxesNode + { + int gamepad_id; + Axis horizontal; + Axis vertical; + float deadzone; + + Vec2 value; + + void init(int gamepad_id, Axis horizontal, Axis vertical, float deadzone); + void update(); + }; + + KeysNode m_keys[BLAH_MAX_VIRTUAL_NODES]; + ButtonsNode m_buttons[BLAH_MAX_VIRTUAL_NODES]; + AxesNode m_axes[BLAH_MAX_VIRTUAL_NODES]; + int m_keys_len = 0; + int m_buttons_len = 0; + int m_axes_len = 0; + + float m_press_buffer = 0; + float m_release_buffer = 0; + float m_repeat_delay = 0; + float m_repeat_interval = 0; + + Vec2 m_value = Vec2(); + Point m_value_i = Point(); + Vec2 m_last_value = Vec2(); + Point m_last_value_i = Point(); + bool m_pressed = false; + bool m_released = false; + + float m_i_deadzone; + float m_last_press_time = 0; + float m_last_release_time = 0; + float m_repeat_press_time = 0; + + public: + VirtualStick(); + VirtualStick(float iDeadzone); + VirtualStick& add_keys(Key left, Key right, Key up, Key down); + VirtualStick& add_buttons(int gamepad_id, Button left, Button right, Button up, Button down); + VirtualStick& add_axes(int gamepad_id, Axis horizontal, Axis vertical, float deadzone); + VirtualStick& repeat(float m_repeat_delay, float m_repeat_interval); + VirtualStick& press_buffer(float duration); + VirtualStick& release_buffer(float duration); + + void update(); + const Vec2& value() const { return m_value; } + const Point& value_i() const { return m_value_i; } + const Vec2& last_value() const { return m_last_value; } + const Point& last_value_i() const { return m_last_value_i; } + bool pressed() const { return m_pressed; } + bool released() const { return m_released; } + void clear_press_buffer() { m_last_press_time = 0; } + void clear_release_buffer() { m_last_release_time = 0; } + }; +} \ No newline at end of file diff --git a/public/blah/log.cpp b/public/blah/log.cpp new file mode 100644 index 0000000..e31a45f --- /dev/null +++ b/public/blah/log.cpp @@ -0,0 +1,60 @@ +#include +#include +#include // for logging methods +#include // for sprintf + +using namespace Blah; + +void Log::print(const char* format, ...) +{ + char msg[BLAH_MAX_MESSAGE]; + va_list ap; + va_start(ap, format); + vsnprintf(msg, sizeof(char) * BLAH_MAX_MESSAGE, format, ap); + va_end(ap); + + if (App::is_running() && App::config()->on_info != nullptr) + { + App::config()->on_info(msg); + } + else + { + printf("%s\n", msg); + } +} + +void Log::warn(const char* format, ...) +{ + char msg[BLAH_MAX_MESSAGE]; + va_list ap; + va_start(ap, format); + vsnprintf(msg, sizeof(char) * BLAH_MAX_MESSAGE, format, ap); + va_end(ap); + + if (App::is_running() && App::config()->on_warn != nullptr) + { + App::config()->on_warn(msg); + } + else + { + printf("\033[01;33mWARN:\033[0m\t%s\n", msg); + } +} + +void Log::error(const char* format, ...) +{ + char msg[BLAH_MAX_MESSAGE]; + va_list ap; + va_start(ap, format); + vsnprintf(msg, sizeof(char) * BLAH_MAX_MESSAGE, format, ap); + va_end(ap); + + if (App::is_running() && App::config()->on_error != nullptr) + { + App::config()->on_error(msg); + } + else + { + printf("\033[1;31mERROR:\033[0m\t%s\n", msg); + } +} \ No newline at end of file diff --git a/public/blah/log.h b/public/blah/log.h new file mode 100644 index 0000000..587c8a3 --- /dev/null +++ b/public/blah/log.h @@ -0,0 +1,34 @@ +#pragma once + +// Error / Abort +#ifdef DEBUG + + #define BLAH_ERROR(message) \ + do { Log::error(message "\n\tin file: %s:%d", __FILE__, __LINE__); abort(); } while(0) + + #define BLAH_ASSERT(condition, message) \ + do { if (!(condition)) { BLAH_ERROR(message); } } while(0) + +#else + + #define BLAH_ERROR(message) \ + Log::error(message "\n\tin file: %s:%d", __FILE__, __LINE__) + + #define BLAH_ASSERT(condition, message) \ + do { } while(false) + +#endif + +#ifndef BLAH_MAX_MESSAGE +#define BLAH_MAX_MESSAGE 1024 +#endif + +namespace Blah +{ + namespace Log + { + void print(const char* info, ...); + void warn(const char* info, ...); + void error(const char* info, ...); + } +} \ No newline at end of file diff --git a/public/blah/math/calc.cpp b/public/blah/math/calc.cpp new file mode 100644 index 0000000..1df36af --- /dev/null +++ b/public/blah/math/calc.cpp @@ -0,0 +1,135 @@ +#include +#include +#include + +using namespace Blah; + +float Calc::rand_float(float min, float maxExc) +{ + return min + rand_float(maxExc - min); +} + +float Calc::rand_float(float maxExc) +{ + return (rand() / (float)RAND_MAX) * maxExc; +} + +int Calc::rand_int(int min, int maxExc) +{ + return min + rand_int(maxExc - min); +} + +int Calc::rand_int(int maxExc) +{ + if (maxExc <= 0) + return 0; + return rand() % maxExc; +} + +int Calc::rand_int() +{ + return rand(); +} + +float Calc::approach(float t, float target, float maxDelta) +{ + return t < target ? min(t + maxDelta, target) : max(t - maxDelta, target); +} + +float Calc::clamp(float t, float min, float max) +{ + return t < min ? min : (t > max ? max : t); +} + +int Calc::clamp_int(int t, int min, int max) +{ + return t < min ? min : (t > max ? max : t); +} + +int Calc::sign(int x) +{ + return (x < 0 ? -1 : (x > 0 ? 1 : 0)); +} + +float Calc::sign(float x) +{ + return (x < 0 ? -1.0f : (x > 0 ? 1.0f : 0.0f)); +} + +int Calc::abs(int x) +{ + return x < 0 ? -x : x; +} + +float Calc::abs(float x) +{ + return x < 0 ? -x : x; +} + +float Calc::floor(float x) +{ + return floorf(x); +} + +float Calc::ceiling(float x) +{ + return ceilf(x); +} + +float Calc::mod(float x, float m) +{ + return x - (int)(x / m) * m; +} + +float Calc::sin(float x) +{ + return sinf(x); +} + +float Calc::cos(float x) +{ + return cosf(x); +} + +float Calc::atan2(float y, float x) +{ + return atan2f(y, x); +} + +float Calc::pow(float x, float n) +{ + return powf(x, n); +} + +float Calc::sqrt(float x) +{ + return sqrtf(x); +} + +float Calc::snap(float val, float interval) +{ + if (val > 0) + return ((int)((val + interval / 2) / interval)) * interval; + else + return ((int)((val - interval / 2) / interval)) * interval; +} + +float Calc::angle_diff(float radians_a, float radians_b) +{ + return mod((radians_b - radians_a) + PI, TAU) - PI; +} + +float Calc::lerp(float a, float b, float t) +{ + return a + (b - a) * t; +} + +bool Calc::is_big_endian() +{ + return (*((short*)"AB") == 0x4243); +} + +bool Calc::is_little_endian() +{ + return (*((short*)"AB") != 0x4243); +} \ No newline at end of file diff --git a/public/blah/math/calc.h b/public/blah/math/calc.h new file mode 100644 index 0000000..6a77e35 --- /dev/null +++ b/public/blah/math/calc.h @@ -0,0 +1,71 @@ +#pragma once +#include + +namespace Blah +{ + namespace Calc + { + constexpr float PI = 3.141592653f; + constexpr float TAU = PI * 2.0f; + constexpr float RIGHT = 0; + constexpr float LEFT = PI; + constexpr float UP = PI / -2; + constexpr float DOWN = PI / 2; + + float rand_float(float min, float maxExc); + + float rand_float(float maxExc); + + int rand_int(int min, int maxExc); + + int rand_int(int maxExc); + + int rand_int(); + + float approach(float t, float target, float maxDelta); + + float clamp(float t, float min, float max); + + int clamp_int(int t, int min, int max); + + int sign(int x); + + float sign(float x); + + int abs(int x); + + float abs(float x); + + template + T min(T a, T b) { return a < b ? a : b; } + + template + T max(T a, T b) { return a > b ? a : b; } + + float floor(float x); + + float ceiling(float x); + + float mod(float x, float m); + + float sin(float x); + + float cos(float x); + + float atan2(float y, float x); + + float pow(float x, float n); + + float sqrt(float x); + + float snap(float val, float interval); + + float angle_diff(float radians_a, float radians_b); + + float lerp(float a, float b, float t); + + bool is_big_endian(); + + bool is_little_endian(); + }; +} diff --git a/public/blah/math/circle.cpp b/public/blah/math/circle.cpp new file mode 100644 index 0000000..e4bda69 --- /dev/null +++ b/public/blah/math/circle.cpp @@ -0,0 +1,11 @@ +#include +#include + +using namespace Blah; + + +void Circle::project(const Vec2& axis, float* min, float* max) const +{ + *min = Vec2::dot(center - axis * radius, axis); + *max = Vec2::dot(center + axis * radius, axis); +} \ No newline at end of file diff --git a/public/blah/math/circle.h b/public/blah/math/circle.h new file mode 100644 index 0000000..95fa1c0 --- /dev/null +++ b/public/blah/math/circle.h @@ -0,0 +1,19 @@ +#pragma once +#include + +namespace Blah +{ + struct Circle + { + Vec2 center; + float radius; + + Circle() + : center(), radius(0) {} + + Circle(Vec2 center, float radius) + : center(center), radius(radius) {} + + void project(const Vec2& axis, float* min, float* max) const; + }; +} \ No newline at end of file diff --git a/public/blah/math/color.cpp b/public/blah/math/color.cpp new file mode 100644 index 0000000..22923a9 --- /dev/null +++ b/public/blah/math/color.cpp @@ -0,0 +1,156 @@ +#include +#include + +using namespace Blah; + +char const hex[16] = { '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', 'A', 'B', 'C', 'D', 'E', 'F' }; +#define LT_HEX_VALUE(n) ((n >= '0' && n <= '9') ? (n - '0') : ((n >= 'A' && n <= 'F') ? (10 + n - 'A') : ((n >= 'a' && n <= 'f') ? (10 + n - 'a') : 0))) + +Color::Color() + : r(0), g(0), b(0), a(0) {} + +Color::Color(int rgb) : + r((uint8_t)((rgb & 0xFF0000) >> 16)), + g((uint8_t)((rgb & 0x00FF00) >> 8)), + b((uint8_t)(rgb & 0x0000FF)), + a(255) {} + +Color::Color(int rgb, float alpha) : + r((int)(((uint8_t)((rgb & 0xFF0000) >> 16)) * alpha)), + g((int)(((uint8_t)((rgb & 0x00FF00) >> 8)) * alpha)), + b((int)(((uint8_t)(rgb & 0x0000FF)) * alpha)), + a((int)(255 * alpha)) {} + +Color::Color(uint8_t r, uint8_t g, uint8_t b) + : r(r), g(g), b(b), a(255) {} + +Color::Color(uint8_t r, uint8_t g, uint8_t b, uint8_t a) + : r(r), g(g), b(b), a(a) {} + +Color::Color(const char* value) : r(0), g(0), b(0), a(255) +{ + int length = 0; + while (value[length] != '\0' && length < 10) + length ++; + + int offset = 0; + if (length > 0 && value[0] == '#') + offset = 1; + else if (length >= 1 && value[0] == '0' && (value[1] == 'x' || value[1] == 'X')) + offset = 2; + + if (length - offset >= 8) + a = (LT_HEX_VALUE(value[offset + 6]) << 4) + LT_HEX_VALUE(value[offset + 7]); + + if (length - offset >= 6) + { + r = (LT_HEX_VALUE(value[offset + 0]) << 4) + LT_HEX_VALUE(value[offset + 1]); + g = (LT_HEX_VALUE(value[offset + 2]) << 4) + LT_HEX_VALUE(value[offset + 3]); + b = (LT_HEX_VALUE(value[offset + 4]) << 4) + LT_HEX_VALUE(value[offset + 5]); + } +} + +void Color::premultiply() +{ + r = r * a / 255; + g = g * a / 255; + b = b * a / 255; +} + +uint32_t Color::to_rgba() const +{ + return + ((uint32_t)r << 24) | + ((uint32_t)g << 16) | + ((uint32_t)b << 8) | + (uint32_t)a; +} + +Vec4 Color::to_vec4() const +{ + return Vec4(r / 255.0f, g / 255.0f, b / 255.0f, a / 255.0f); +} + +void Color::to_hex_rgba(char* buffer) const +{ + buffer[0] = hex[(r & 0xF0) >> 4]; + buffer[1] = hex[(r & 0x0F) >> 0]; + buffer[2] = hex[(g & 0xF0) >> 4]; + buffer[3] = hex[(g & 0x0F) >> 0]; + buffer[4] = hex[(b & 0xF0) >> 4]; + buffer[5] = hex[(b & 0x0F) >> 0]; + buffer[6] = hex[(a & 0xF0) >> 4]; + buffer[7] = hex[(a & 0x0F) >> 0]; +} + +void Color::to_hex_rgb(char* buffer) const +{ + buffer[0] = hex[(r & 0xF0) >> 4]; + buffer[1] = hex[(r & 0x0F) >> 0]; + buffer[2] = hex[(g & 0xF0) >> 4]; + buffer[3] = hex[(g & 0x0F) >> 0]; + buffer[4] = hex[(b & 0xF0) >> 4]; + buffer[5] = hex[(b & 0x0F) >> 0]; +} + +Color Color::from_rgba(uint32_t value) +{ + return + { + (uint8_t)((value & 0xFF000000) >> 24), + (uint8_t)((value & 0x00FF0000) >> 16), + (uint8_t)((value & 0x0000FF00) >> 8), + (uint8_t)((value & 0x000000FF)) + }; +} + +Color Color::from_rgb(uint32_t value) +{ + return + { + (uint8_t)((value & 0xFF0000) >> 16), + (uint8_t)((value & 0x00FF00) >> 8), + (uint8_t)((value & 0x0000FF)) + }; +} + +Color Color::lerp(Color a, Color b, float amount) +{ + if (amount < 0) amount = 0; + if (amount > 1) amount = 1; + + return Color( + (uint8_t)(a.r + (b.r - a.r) * amount), + (uint8_t)(a.g + (b.g - a.g) * amount), + (uint8_t)(a.b + (b.b - a.b) * amount), + (uint8_t)(a.a + (b.a - a.a) * amount) + ); +} + +Color Color::operator*(float multiply) const +{ + return Color( + (int)(r * multiply), + (int)(g * multiply), + (int)(b * multiply), + (int)(a * multiply)); +} + +Color& Color::operator= (const int rgb) +{ + r = (uint8_t)((rgb & 0xFF0000) >> 16); + g = (uint8_t)((rgb & 0x00FF00) >> 8); + b = (uint8_t)(rgb & 0x0000FF); + a = 255; + return *this; +} + +const Color Color::transparent = Color(0, 0, 0, 0); +const Color Color::white = Color(255, 255, 255, 255); +const Color Color::black = Color(0, 0, 0, 255); +const Color Color::red = Color(255, 0, 0, 255); +const Color Color::green = Color(0, 255, 0, 255); +const Color Color::blue = Color(0, 0, 255, 255); +const Color Color::yellow = Color(255, 255, 0, 255); +const Color Color::purple = Color(255, 0, 255, 255); +const Color Color::teal = Color(0, 255, 255, 255); \ No newline at end of file diff --git a/public/blah/math/color.h b/public/blah/math/color.h new file mode 100644 index 0000000..0b99316 --- /dev/null +++ b/public/blah/math/color.h @@ -0,0 +1,64 @@ +#pragma once +#include + +namespace Blah +{ + struct Vec4; + + struct Color + { + uint8_t r; + uint8_t g; + uint8_t b; + uint8_t a; + + Color(); + Color(int rgb); + Color(int rgb, float alpha); + Color(uint8_t r, uint8_t g, uint8_t b); + Color(uint8_t r, uint8_t g, uint8_t b, uint8_t a); + + // Parses a Hex string in the format of "#00000000" or "0x00000000" or "00000000" + Color(const char* hexCstr); + + // Premultiplies the Color + void premultiply(); + + // Sets a Hex string to the given buffer, in the format of RRGGBBAA + // The buffer must be at least 8 bytes long + void to_hex_rgba(char* buffer) const; + + // Sets a Hex string to the given buffer, in the format of RRGGBB + // The buffer must be at least 6 bytes long + void to_hex_rgb(char* buffer) const; + + uint32_t to_rgba() const; + + Vec4 to_vec4() const; + + // Returns a RGBA Color representation of the integer value + static Color from_rgba(uint32_t value); + + // Returns a RGB Color representation of the integer value, with Alpha = 255 + static Color from_rgb(uint32_t value); + + static Color lerp(Color a, Color b, float amount); + + // Mutliplties the Color + Color operator*(float multiply) const; + + // assignment from int + Color& operator= (const int rgb); + + static const Color transparent; + static const Color white; + static const Color black; + static const Color red; + static const Color green; + static const Color blue; + static const Color yellow; + static const Color orange; + static const Color purple; + static const Color teal; + }; +} diff --git a/public/blah/math/ease.h b/public/blah/math/ease.h new file mode 100644 index 0000000..457db86 --- /dev/null +++ b/public/blah/math/ease.h @@ -0,0 +1,336 @@ +#pragma once +#include +#include + +namespace Blah +{ + using Easer = float (*)(float); + + enum class Easers + { + QuadIn, QuadOut, QuadInOut, + CubeIn, CubeOut, CubeInOut, + QuartIn, QuartOut, QuartInOut, + QuintIn, QuintOut, QuintInOut, + SineIn, SineOut, SineInOut, + CircIn, CircOut, CircInOut, + ExpIn, ExpOut, ExpInOut, + ElasticIn, ElasticOut, ElasticInOut, + BackIn, BackOut, BackInOut, + BounceIn, BounceOut, BounceInOut, + _Count + }; + + namespace Ease + { + /* + Adapted from functions here: https://github.com/warrenm/AHEasing/blob/master/AHEasing/easing.c + For previews go here: https://easings.net/ + */ + + float linear(float t) + { + return t; + } + + float quad_in(float t) + { + return t * t; + } + + float quad_out(float t) + { + return -(t * (t - 2)); + } + + float quad_in_out(float t) + { + if (t < 0.5f) + return 2 * t * t; + else + { + return (-2 * t * t) + (4 * t) - 1; + } + } + + float cube_in(float t) + { + return t * t * t; + } + + float cube_out(float t) + { + float f = (t - 1); + return f * f * f + 1; + } + + float cube_in_out(float t) + { + if (t < 0.5f) + return 4 * t * t * t; + else + { + float f = ((2 * t) - 2); + return 0.5f * f * f * f + 1; + } + } + + float quart_in(float t) + { + return t * t * t * t; + } + + float quart_out(float t) + { + float f = (t - 1); + return f * f * f * (1 - t) + 1; + } + + float quart_in_out(float t) + { + if (t < 0.5f) + return 8 * t * t * t * t; + else + { + float f = (t - 1); + return -8 * f * f * f * f + 1; + } + } + + float quint_in(float t) + { + return t * t * t * t * t; + } + + float quint_out(float t) + { + float f = (t - 1); + return f * f * f * f * f + 1; + } + + float quint_in_out(float t) + { + if (t < 0.5f) + return 16 * t * t * t * t * t; + else + { + float f = ((2 * t) - 2); + return 0.5f * f * f * f * f * f + 1; + } + } + + float sine_in(float t) + { + return Calc::sin((t - 1) * Calc::PI * 0.5f) + 1; + } + + float sine_out(float t) + { + return Calc::sin(t * (Calc::PI * 0.5f)); + } + + float sine_in_out(float t) + { + return 0.5f * (1 - Calc::cos(t * Calc::PI)); + } + + float circ_in(float t) + { + return 1 - Calc::sqrt(1 - (t * t)); + } + + float circ_out(float t) + { + return Calc::sqrt((2 - t) * t); + } + + float circ_in_out(float t) + { + if (t < 0.5f) + return 0.5f * (1 - Calc::sqrt(1 - 4 * (t * t))); + else + return 0.5f * (Calc::sqrt(-((2 * t) - 3) * ((2 * t) - 1)) + 1); + } + + float exp_in(float t) + { + return (t == 0) ? 0 : Calc::pow(2, 10 * (t - 1)); + } + + float exp_out(float t) + { + return (t == 1) ? 1 : 1 - Calc::pow(2, -10 * t); + } + + float exp_in_out(float t) + { + if (t == 0 || t == 1) + return t; + + if (t < 0.5f) + return 0.5f * Calc::pow(2, (20 * t) - 10); + else + return -0.5f * Calc::pow(2, (-20 * t) + 10) + 1; + } + + float elastic_in(float t) + { + return Calc::sin(13 * (Calc::PI * 0.5f) * t) * Calc::pow(2, 10 * (t - 1)); + } + + float elastic_out(float t) + { + return Calc::sin(-13 * (Calc::PI * 0.5f) * (t + 1)) * Calc::pow(2, -10 * t) + 1; + } + + float elastic_in_out(float t) + { + if (t < 0.5f) + return 0.5f * Calc::sin(13 * (Calc::PI * 0.5f) * (2 * t)) * Calc::pow(2, 10 * ((2 * t) - 1)); + else + return 0.5f * (Calc::sin(-13 * (Calc::PI * 0.5f) * ((2 * t - 1) + 1)) * Calc::pow(2, -10 * (2 * t - 1)) + 2); + } + + float back_in(float t) + { + return t * t * t - t * Calc::sin(t * Calc::PI); + } + + float back_out(float t) + { + float f = (1 - t); + return 1 - (f * f * f - f * Calc::sin(f * Calc::PI)); + } + + float back_in_out(float t) + { + if (t < 0.5f) + { + float f = 2 * t; + return 0.5f * (f * f * f - f * Calc::sin(f * Calc::PI)); + } + else + { + float f = (1 - (2 * t - 1)); + return 0.5f * (1 - (f * f * f - f * Calc::sin(f * Calc::PI))) + 0.5f; + } + } + + float bounce_out(float t) + { + if (t < 4 / 11.0f) + return (121 * t * t) / 16.0f; + else if (t < 8 / 11.0f) + return (363 / 40.0f * t * t) - (99 / 10.0f * t) + 17 / 5.0f; + else if (t < 9 / 10.0f) + return (4356 / 361.0f * t * t) - (35442 / 1805.0f * t) + 16061 / 1805.0f; + else + return (54 / 5.0f * t * t) - (513 / 25.0f * t) + 268 / 25.0f; + } + + float bounce_in(float t) + { + return 1 - bounce_out(1 - t); + } + + float bounce_in_out(float t) + { + if (t < 0.5f) + return 0.5f * bounce_in(t * 2); + else + return 0.5f * bounce_out(t * 2 - 1) + 0.5f; + } + + Easer get(Easers e) + { + switch (e) + { + case Easers::CubeIn: return &cube_in; + case Easers::CubeOut: return &cube_out; + case Easers::CubeInOut: return &cube_in_out; + + case Easers::QuadIn: return &quad_in; + case Easers::QuadOut: return &quad_out; + case Easers::QuadInOut: return &quad_in_out; + + case Easers::QuartIn: return &quart_in; + case Easers::QuartOut: return &quart_out; + case Easers::QuartInOut: return &quart_in_out; + + case Easers::QuintIn: return &quint_in; + case Easers::QuintOut: return &quint_out; + case Easers::QuintInOut: return &quint_in_out; + + case Easers::SineIn: return &sine_in; + case Easers::SineOut: return &sine_out; + case Easers::SineInOut: return &sine_in_out; + + case Easers::CircIn: return &circ_in; + case Easers::CircOut: return &circ_out; + case Easers::CircInOut: return &circ_in_out; + + case Easers::ExpIn: return &exp_in; + case Easers::ExpOut: return &exp_out; + case Easers::ExpInOut: return &exp_in_out; + + case Easers::ElasticIn: return &elastic_in; + case Easers::ElasticOut: return &elastic_out; + case Easers::ElasticInOut: return &elastic_in_out; + + case Easers::BackIn: return &back_in; + case Easers::BackOut: return &back_out; + case Easers::BackInOut: return &back_in_out; + + case Easers::BounceIn: return &bounce_in; + case Easers::BounceOut: return &bounce_out; + case Easers::BounceInOut: return &bounce_in_out; + case Easers::_Count: + break; + } + + BLAH_ERROR("Invalid Easer Type"); + return nullptr; + } + + const char* name(Easers e) + { + switch (e) + { + case Easers::CubeIn: return "CubeIn"; + case Easers::CubeOut: return "CubeOut"; + case Easers::CubeInOut: return "CubeInOut"; + case Easers::QuadIn: return "QuadIn"; + case Easers::QuadOut: return "QuadOut"; + case Easers::QuadInOut: return "QuadInOut"; + case Easers::QuartIn: return "QuartIn"; + case Easers::QuartOut: return "QuartOut"; + case Easers::QuartInOut: return "QuartInOut"; + case Easers::QuintIn: return "QuintIn"; + case Easers::QuintOut: return "QuintOut"; + case Easers::QuintInOut: return "QuintInOut"; + case Easers::SineIn: return "SineIn"; + case Easers::SineOut: return "SineOut"; + case Easers::SineInOut: return "SineInOut"; + case Easers::CircIn: return "CircIn"; + case Easers::CircOut: return "CircOut"; + case Easers::CircInOut: return "CircInOut"; + case Easers::ExpIn: return "ExpIn"; + case Easers::ExpOut: return "ExpOut"; + case Easers::ExpInOut: return "ExpInOut"; + case Easers::ElasticIn: return "ElasticIn"; + case Easers::ElasticOut: return "ElasticOut"; + case Easers::ElasticInOut: return "ElasticInOut"; + case Easers::BackIn: return "BackIn"; + case Easers::BackOut: return "BackOut"; + case Easers::BackInOut: return "BackInOut"; + case Easers::BounceIn: return "BounceIn"; + case Easers::BounceOut: return "BounceOut"; + case Easers::BounceInOut: return "BounceInOut"; + case Easers::_Count: + break; + } + + return ""; + } + } +} diff --git a/public/blah/math/line.cpp b/public/blah/math/line.cpp new file mode 100644 index 0000000..fce49bb --- /dev/null +++ b/public/blah/math/line.cpp @@ -0,0 +1,157 @@ +#include +#include +#include +#include + +using namespace Blah; + +Line::Line(float x0, float y0, float x1, float y1) +{ + a.x = x0; + a.y = y0; + b.x = x1; + b.y = y1; +} + +Line::Line(const Vec2& start, const Vec2& end) + : a(start), b(end) {} + +Rect Line::bounds() const +{ + Vec2 pos = Vec2(Calc::min(a.x, b.x), Calc::min(a.y, b.y)); + + return Rect( + pos, + Vec2(Calc::max(a.x, b.x) - pos.x, Calc::max(a.y, b.y) - pos.y) + ); +} + +Vec2 Line::closest_point(const Vec2& pt) const +{ + Vec2 v = b - a; + Vec2 w = pt - a; + float t = Vec2::dot(w, v) / Vec2::dot(v, v); + t = Calc::clamp(t, 0, 1); + + return v * t + a; +} + +bool Line::intersects(const Rect& rect) const +{ + char ca = rect.get_sector(a); + char cb = rect.get_sector(b); + + if (ca == cb || (ca & cb) != 0) + return false; + + char both = ca | cb; + + // top + if ((both & 0b0100) != 0 && intersects(rect.top_line())) + return true; + + // bottom + if ((both & 0b1000) != 0 && intersects(rect.bottom_line())) + return true; + + // left + if ((both & 0b0001) != 0 && intersects(rect.left_line())) + return true; + + // right + if ((both & 0b0010) != 0 && intersects(rect.right_line())) + return true; + + return false; +} + +bool Line::intersects(const Rect& rect, Vec2* out_intersection_point) const +{ + char ca = rect.get_sector(a); + char cb = rect.get_sector(b); + + if (ca == cb || (ca & cb) != 0) + return false; + + char both = ca | cb; + + // top + if ((both & 0b0100) != 0 && intersects(rect.top_line(), out_intersection_point)) + return true; + + // bottom + if ((both & 0b1000) != 0 && intersects(rect.bottom_line(), out_intersection_point)) + return true; + + // left + if ((both & 0b0001) != 0 && intersects(rect.left_line(), out_intersection_point)) + return true; + + // right + if ((both & 0b0010) != 0 && intersects(rect.right_line(), out_intersection_point)) + return true; + + return false; +} + +bool Line::intersects(const Line& line) const +{ + Vec2 e = b - a; + Vec2 d = line.b - line.a; + float e_dot_d_perp = e.x * d.y - e.y * d.x; + + // if e dot d == 0, it means the lines are parallel + // so have infinite intersection points + if (e_dot_d_perp < 0.0001 && e_dot_d_perp > -0.0001) + return false; + + Vec2 c = line.a - a; + float t = (c.x * d.y - c.y * d.x) / e_dot_d_perp; + if (t < 0 || t > 1) + return false; + + float u = (c.x * e.y - c.y * e.x) / e_dot_d_perp; + if (u < 0 || u > 1) + return false; + + return true; +} + +bool Line::intersects(const Line& line, Vec2* intersection_point) const +{ + Vec2 e = b - a; + Vec2 d = line.b - line.a; + float e_dot_d_perp = e.x * d.y - e.y * d.x; + + // if e dot d == 0, it means the lines are parallel + // so have infinite intersection points + if (e_dot_d_perp < 0.0001 && e_dot_d_perp > -0.0001) + return false; + + Vec2 c = line.a - a; + float t = (c.x * d.y - c.y * d.x) / e_dot_d_perp; + if (t < 0 || t > 1) + return false; + + float u = (c.x * e.y - c.y * e.x) / e_dot_d_perp; + if (u < 0 || u > 1) + return false; + + Vec2 i = (e * t) + a; + intersection_point->x = i.x; + intersection_point->y = i.y; + return true; +} + +void Line::project(const Vec2& axis, float* min, float* max) const +{ + float dot = a.x * axis.x + a.y * axis.y; + *min = dot; + *max = dot; + dot = b.x * axis.x + b.y * axis.y; + *min = Calc::min(dot, *min); + *max = Calc::max(dot, *max); +} + +Line Line::operator +(const Vec2& rhs) const { return Line(a + rhs, b + rhs); } +Line Line::operator -(const Vec2& rhs) const { return Line(a - rhs, b - rhs); } diff --git a/public/blah/math/line.h b/public/blah/math/line.h new file mode 100644 index 0000000..14e2c50 --- /dev/null +++ b/public/blah/math/line.h @@ -0,0 +1,30 @@ +#pragma once +#include + +namespace Blah +{ + struct Rect; + + struct Line + { + Vec2 a; + Vec2 b; + + Line() {} + Line(float x0, float y0, float x1, float y1); + Line(const Vec2& start, const Vec2& end); + + Rect bounds() const; + + Vec2 closest_point(const Vec2& pt) const; + bool intersects(const Rect& rect) const; + bool intersects(const Rect& rect, Vec2* out_intersection_point) const; + bool intersects(const Line& line) const; + bool intersects(const Line& line, Vec2* out_intersection_point) const; + + void project(const Vec2& axis, float* min, float* max) const; + + Line operator +(const Vec2& rhs) const; + Line operator -(const Vec2& rhs) const; + }; +} \ No newline at end of file diff --git a/public/blah/math/mat3x2.cpp b/public/blah/math/mat3x2.cpp new file mode 100644 index 0000000..a9a0a1b --- /dev/null +++ b/public/blah/math/mat3x2.cpp @@ -0,0 +1,187 @@ +#include +#include +#include +#include + +using namespace Blah; + +Mat3x2::Mat3x2() + : m11(0), m12(0), m21(0), m22(0), m31(0), m32(0) {} + +Mat3x2::Mat3x2(float m11, float m12, float m21, float m22, float m31, float m32) + : m11(m11), m12(m12), m21(m21), m22(m22), m31(m31), m32(m32) {} + +Mat3x2 Mat3x2::invert() const +{ + auto det = (m11 * m22) - (m21 * m12); + auto invDet = 1.0f / det; + + return Mat3x2 + { + m22 * invDet, + -m12 * invDet, + -m21 * invDet, + m11 * invDet, + (m21 * m32 - m31 * m22) * invDet, + (m31 * m12 - m11 * m32) * invDet + }; +} + +float Mat3x2::scaling_factor() const +{ + return (float)sqrt((double)m11 * m11 + (double)m12 * m12); +} + +const Mat3x2 Mat3x2::identity = Mat3x2(1, 0, 0, 1, 0, 0); + +Mat3x2 Mat3x2::operator *(const Mat3x2& rhs) const { return multiply(*this, rhs); } +Mat3x2 Mat3x2::operator +(const Mat3x2& rhs) const { return add(*this, rhs); } +Mat3x2 Mat3x2::operator -(const Mat3x2& rhs) const { return subtract(*this, rhs); } + +Mat3x2& Mat3x2::operator*=(const Mat3x2& rhs) +{ + *this = multiply(*this, rhs); + return *this; +} + +bool Mat3x2::operator ==(const Mat3x2& rhs) +{ + return memcmp(this, &rhs, sizeof(Mat3x2)); +} + +bool Mat3x2::operator !=(const Mat3x2& rhs) +{ + return !(*this == rhs); +} + +Mat3x2 Mat3x2::create_translation(const Vec2& Vec2) +{ + return create_translation(Vec2.x, Vec2.y); +} + +Mat3x2 Mat3x2::create_translation(float x, float y) +{ + return Mat3x2(1, 0, 0, 1, x, y); +} + +Mat3x2 Mat3x2::create_scale(float scale) +{ + return create_scale(scale, scale); +} + +Mat3x2 Mat3x2::create_scale(Vec2 Vec2) +{ + return create_scale(Vec2.x, Vec2.y); +} + +Mat3x2 Mat3x2::create_scale(float x, float y) +{ + return Mat3x2(x, 0, 0, y, 0, 0); +} + +Mat3x2 Mat3x2::create_scale(float scale, Vec2 centerPoint) +{ + Mat3x2 result; + + float tx = centerPoint.x * (1 - scale); + float ty = centerPoint.y * (1 - scale); + + result.m11 = scale; + result.m12 = 0.0f; + result.m21 = 0.0f; + result.m22 = scale; + result.m31 = tx; + result.m32 = ty; + + return result; +} + +Mat3x2 Mat3x2::create_scale(Vec2 scale, Vec2 centerPoint) +{ + Mat3x2 result; + + float tx = centerPoint.x * (1 - scale.x); + float ty = centerPoint.y * (1 - scale.y); + + result.m11 = scale.x; + result.m12 = 0.0f; + result.m21 = 0.0f; + result.m22 = scale.y; + result.m31 = tx; + result.m32 = ty; + + return result; +} + +Mat3x2 Mat3x2::create_scale(float scaleX, float scaleY, Vec2 centerPoint) +{ + Mat3x2 result; + + float tx = centerPoint.x * (1 - scaleX); + float ty = centerPoint.y * (1 - scaleY); + + result.m11 = scaleX; + result.m12 = 0.0f; + result.m21 = 0.0f; + result.m22 = scaleY; + result.m31 = tx; + result.m32 = ty; + + return result; +} + +Mat3x2 Mat3x2::create_rotation(float radians) +{ + float c = cosf(radians); + float s = sinf(radians); + + return Mat3x2(c, s, -s, c, 0, 0); +} + +Mat3x2 Mat3x2::create_transform(const Vec2& position, const Vec2& origin, const Vec2& scale, float rotation) +{ + Mat3x2 matrix = identity; + + if (origin.x != 0 || origin.y != 0) + matrix = create_translation(-origin.x, -origin.y); + if (scale.x != 1 || scale.y != 1) + matrix = matrix * create_scale(scale); + if (rotation != 0) + matrix = matrix * create_rotation(rotation); + if (position.x != 0 || position.y != 0) + matrix = matrix * create_translation(position); + + return matrix; +} + +Mat3x2 Mat3x2::add(const Mat3x2& a, const Mat3x2& b) +{ + return Mat3x2( + a.m11 + b.m11, + a.m12 + b.m12, + a.m21 + b.m21, + a.m22 + b.m22, + a.m31 + b.m31, + a.m21 + b.m32); +} + +Mat3x2 Mat3x2::subtract(const Mat3x2& a, const Mat3x2& b) +{ + return Mat3x2( + a.m11 - b.m11, + a.m12 - b.m12, + a.m21 - b.m21, + a.m22 - b.m22, + a.m31 - b.m31, + a.m21 - b.m32); +} + +Mat3x2 Mat3x2::multiply(const Mat3x2& a, const Mat3x2& b) +{ + return Mat3x2(a.m11 * b.m11 + a.m12 * b.m21, + a.m11 * b.m12 + a.m12 * b.m22, + a.m21 * b.m11 + a.m22 * b.m21, + a.m21 * b.m12 + a.m22 * b.m22, + a.m31 * b.m11 + a.m32 * b.m21 + b.m31, + a.m31 * b.m12 + a.m32 * b.m22 + b.m32); +} \ No newline at end of file diff --git a/public/blah/math/mat3x2.h b/public/blah/math/mat3x2.h new file mode 100644 index 0000000..7625ddb --- /dev/null +++ b/public/blah/math/mat3x2.h @@ -0,0 +1,47 @@ +#pragma once + +namespace Blah +{ + struct Vec2; + + struct Mat3x2 + { + float m11; + float m12; + float m21; + float m22; + float m31; + float m32; + + Mat3x2(); + Mat3x2(float m11, float m12, float m21, float m22, float m31, float m32); + + Mat3x2 invert() const; + float scaling_factor() const; + + Mat3x2 operator *(const Mat3x2& rhs) const; + Mat3x2 operator +(const Mat3x2& rhs) const; + Mat3x2 operator -(const Mat3x2& rhs) const; + Mat3x2& operator *=(const Mat3x2& rhs); + bool operator==(const Mat3x2& rhs); + bool operator!=(const Mat3x2& rhs); + + static const Mat3x2 identity; + + static Mat3x2 create_translation(const Vec2& position); + static Mat3x2 create_translation(float x, float y); + static Mat3x2 create_scale(float scale); + static Mat3x2 create_scale(Vec2 scale); + static Mat3x2 create_scale(float x, float y); + static Mat3x2 create_scale(float scale, Vec2 centerPoint); + static Mat3x2 create_scale(Vec2 scale, Vec2 centerPoint); + static Mat3x2 create_scale(float scaleX, float scaleY, Vec2 centerPoint); + static Mat3x2 create_rotation(float radians); + static Mat3x2 create_transform(const Vec2& position, const Vec2& origin, const Vec2& scale, float rotation); + + static Mat3x2 add(const Mat3x2& a, const Mat3x2& b); + static Mat3x2 subtract(const Mat3x2& a, const Mat3x2& b); + static Mat3x2 multiply(const Mat3x2& a, const Mat3x2& b); + + }; +} \ No newline at end of file diff --git a/public/blah/math/mat4x4.cpp b/public/blah/math/mat4x4.cpp new file mode 100644 index 0000000..121aafc --- /dev/null +++ b/public/blah/math/mat4x4.cpp @@ -0,0 +1,109 @@ +#include + +using namespace Blah; + +Mat4x4::Mat4x4() : + m11(0.0f), m12(0.0f), m13(0.0f), m14(0.0f), + m21(0.0f), m22(0.0f), m23(0.0f), m24(0.0f), + m31(0.0f), m32(0.0f), m33(0.0f), m34(0.0f), + m41(0.0f), m42(0.0f), m43(0.0f), m44(0.0f) {} + +Mat4x4::Mat4x4( + float m11, float m12, float m13, float m14, + float m21, float m22, float m23, float m24, + float m31, float m32, float m33, float m34, + float m41, float m42, float m43, float m44) : + m11(m11), m12(m12), m13(m13), m14(m14), + m21(m21), m22(m22), m23(m23), m24(m24), + m31(m31), m32(m32), m33(m33), m34(m34), + m41(m41), m42(m42), m43(m43), m44(m44) {} + +const Mat4x4 Mat4x4::Identity = Mat4x4( + 1.0f, 0.0f, 0.0f, 0.0f, + 0.0f, 1.0f, 0.0f, 0.0f, + 0.0f, 0.0f, 1.0f, 0.0f, + 0.0f, 0.0f, 0.0f, 1.0f); + +Mat4x4 Mat4x4::CreateOrthographic(float width, float height, float zNearPlane, float zFarPlane) +{ + Mat4x4 result = Identity; + + result.m11 = 2.0f / width; + result.m12 = result.m13 = result.m14 = 0.0f; + result.m22 = -2.0f / height; + result.m21 = result.m23 = result.m24 = 0.0f; + result.m33 = 1.0f / (zNearPlane - zFarPlane); + result.m31 = result.m32 = result.m34 = 0.0f; + result.m41 = result.m42 = 0.0f; + result.m43 = zNearPlane / (zNearPlane - zFarPlane); + result.m44 = 1.0f; + + return result; +} + +Mat4x4 Mat4x4::CreateOrthographicOffCenter(float left, float right, float bottom, float top, float zNearPlane, float zFarPlane) +{ + Mat4x4 result = Identity; + + result.m11 = 2.0f / (right - left); + result.m12 = result.m13 = result.m14 = 0.0f; + result.m22 = 2.0f / (top - bottom); + result.m21 = result.m23 = result.m24 = 0.0f; + result.m33 = 1.0f / (zNearPlane - zFarPlane); + result.m31 = result.m32 = result.m34 = 0.0f; + result.m41 = (left + right) / (left - right); + result.m42 = (top + bottom) / (bottom - top); + result.m43 = zNearPlane / (zNearPlane - zFarPlane); + result.m44 = 1.0f; + + return result; +} + +Mat4x4 Mat4x4::CreateTranslation(float xPosition, float yPosition, float zPosition) +{ + Mat4x4 result = Identity; + + result.m41 = xPosition; + result.m42 = yPosition; + result.m43 = zPosition; + + return result; +} + +Mat4x4 Mat4x4::CreateScale(float xScale, float yScale, float zScale) +{ + Mat4x4 result = Identity; + + result.m11 = xScale; + result.m22 = yScale; + result.m33 = zScale; + + return result; +} + +Mat4x4 Mat4x4::operator*(const Mat4x4& rhs) +{ + Mat4x4 m; + + m.m11 = m11 * rhs.m11 + m12 * rhs.m21 + m13 * rhs.m31 + m14 * rhs.m41; + m.m12 = m11 * rhs.m12 + m12 * rhs.m22 + m13 * rhs.m32 + m14 * rhs.m42; + m.m13 = m11 * rhs.m13 + m12 * rhs.m23 + m13 * rhs.m33 + m14 * rhs.m43; + m.m14 = m11 * rhs.m14 + m12 * rhs.m24 + m13 * rhs.m34 + m14 * rhs.m44; + + m.m21 = m21 * rhs.m11 + m22 * rhs.m21 + m23 * rhs.m31 + m24 * rhs.m41; + m.m22 = m21 * rhs.m12 + m22 * rhs.m22 + m23 * rhs.m32 + m24 * rhs.m42; + m.m23 = m21 * rhs.m13 + m22 * rhs.m23 + m23 * rhs.m33 + m24 * rhs.m43; + m.m24 = m21 * rhs.m14 + m22 * rhs.m24 + m23 * rhs.m34 + m24 * rhs.m44; + + m.m31 = m31 * rhs.m11 + m32 * rhs.m21 + m33 * rhs.m31 + m34 * rhs.m41; + m.m32 = m31 * rhs.m12 + m32 * rhs.m22 + m33 * rhs.m32 + m34 * rhs.m42; + m.m33 = m31 * rhs.m13 + m32 * rhs.m23 + m33 * rhs.m33 + m34 * rhs.m43; + m.m34 = m31 * rhs.m14 + m32 * rhs.m24 + m33 * rhs.m34 + m34 * rhs.m44; + + m.m41 = m41 * rhs.m11 + m42 * rhs.m21 + m43 * rhs.m31 + m44 * rhs.m41; + m.m42 = m41 * rhs.m12 + m42 * rhs.m22 + m43 * rhs.m32 + m44 * rhs.m42; + m.m43 = m41 * rhs.m13 + m42 * rhs.m23 + m43 * rhs.m33 + m44 * rhs.m43; + m.m44 = m41 * rhs.m14 + m42 * rhs.m24 + m43 * rhs.m34 + m44 * rhs.m44; + + return m; +} \ No newline at end of file diff --git a/public/blah/math/mat4x4.h b/public/blah/math/mat4x4.h new file mode 100644 index 0000000..5ff0e87 --- /dev/null +++ b/public/blah/math/mat4x4.h @@ -0,0 +1,44 @@ +#pragma once + +namespace Blah +{ + struct Mat4x4 + { + float m11; + float m12; + float m13; + float m14; + + float m21; + float m22; + float m23; + float m24; + + float m31; + float m32; + float m33; + float m34; + + float m41; + float m42; + float m43; + float m44; + + Mat4x4(); + Mat4x4( + float m11, float m12, float m13, float m14, + float m21, float m22, float m23, float m24, + float m31, float m32, float m33, float m34, + float m41, float m42, float m43, float m44); + + + static const Mat4x4 Identity; + + static Mat4x4 CreateOrthographic(float width, float height, float zNearPlane, float zFarPlane); + static Mat4x4 CreateOrthographicOffCenter(float left, float right, float bottom, float top, float zNearPlane, float zFarPlane); + static Mat4x4 CreateTranslation(float xPosition, float yPosition, float zPosition); + static Mat4x4 CreateScale(float xScale, float yScale, float zScale); + + Mat4x4 operator* (const Mat4x4& rhs); + }; +} \ No newline at end of file diff --git a/public/blah/math/point.cpp b/public/blah/math/point.cpp new file mode 100644 index 0000000..f0e9a99 --- /dev/null +++ b/public/blah/math/point.cpp @@ -0,0 +1,51 @@ +#include +#include +#include + +using namespace Blah; + +Point::Point() +{ + x = y = 0; +} + +Point::Point(int px, int py) +{ + x = px; + y = py; +} + +Point Point::operator +(const Point rhs) const { return Point(x + rhs.x, y + rhs.y); } +Point Point::operator -(const Point rhs) const { return Point(x - rhs.x, y - rhs.y); } +Point Point::operator /(const int rhs) const { return Point(x / rhs, y / rhs); } +Point Point::operator *(const int rhs) const { return Point(x * rhs, y * rhs); } +Point Point::operator-() const { return Point(-x, -y); } + +Point& Point::operator +=(const Point& rhs) { x += rhs.x; y += rhs.y; return *this; } +Point& Point::operator -=(const Point& rhs) { x -= rhs.x; y -= rhs.y; return *this; } +Point& Point::operator /=(const Point& rhs) { x /= rhs.x; y /= rhs.y; return *this; } +Point& Point::operator *=(const Point& rhs) { x *= rhs.x; y *= rhs.y; return *this; } +Point& Point::operator /=(int rhs) { x /= rhs; y /= rhs; return *this; } +Point& Point::operator *=(int rhs) { x *= rhs; y *= rhs; return *this; } + +bool Point::operator ==(const Point& rhs) { return x == rhs.x && y == rhs.y; } +bool Point::operator !=(const Point& rhs) { return x != rhs.x || y != rhs.y; } + +float Point::length() const +{ + return sqrtf((float)(x * x + y * y)); +} + +int Point::length_squared() const +{ + return x * x + y * y; +} + +const Point Point::unitX = Point(1, 0); +const Point Point::unitY = Point(0, 1); +const Point Point::right = Point(1, 0); +const Point Point::up = Point(0, -1); +const Point Point::down = Point(0, 1); +const Point Point::left = Point(-1, 0); +const Point Point::zero = Point(0, 0); +const Point Point::one = Point(1, 1); diff --git a/public/blah/math/point.h b/public/blah/math/point.h new file mode 100644 index 0000000..011603a --- /dev/null +++ b/public/blah/math/point.h @@ -0,0 +1,41 @@ +#pragma once + +namespace Blah +{ + struct Point + { + int x; + int y; + + Point(); + Point(int px, int py); + + Point operator +(const Point rhs) const; + Point operator -(const Point rhs) const; + Point operator /(const int rhs) const; + Point operator *(const int rhs) const; + Point operator -() const; + + Point& operator +=(const Point& rhs); + Point& operator -=(const Point& rhs); + Point& operator /=(const Point& rhs); + Point& operator *=(const Point& rhs); + Point& operator /=(int rhs); + Point& operator *=(int rhs); + + bool operator ==(const Point& rhs); + bool operator !=(const Point& rhs); + + float length() const; + int length_squared() const; + + static const Point unitX; + static const Point unitY; + static const Point right; + static const Point up; + static const Point down; + static const Point left; + static const Point zero; + static const Point one; + }; +} \ No newline at end of file diff --git a/public/blah/math/quad.cpp b/public/blah/math/quad.cpp new file mode 100644 index 0000000..9330719 --- /dev/null +++ b/public/blah/math/quad.cpp @@ -0,0 +1,21 @@ +#include +#include + +namespace Blah +{ + void Quad::project(const Vec2& axis, float* min, float* max) const + { + float dot = Vec2::dot(a, axis); + *min = dot; + *max = dot; + dot = Vec2::dot(b, axis); + *min = Calc::min(dot, *min); + *max = Calc::max(dot, *max); + dot = Vec2::dot(c, axis); + *min = Calc::min(dot, *min); + *max = Calc::max(dot, *max); + dot = Vec2::dot(d, axis); + *min = Calc::min(dot, *min); + *max = Calc::max(dot, *max); + } +} \ No newline at end of file diff --git a/public/blah/math/quad.h b/public/blah/math/quad.h new file mode 100644 index 0000000..12d1ea1 --- /dev/null +++ b/public/blah/math/quad.h @@ -0,0 +1,19 @@ +#pragma once +#include + +namespace Blah +{ + struct Quad + { + Vec2 a; + Vec2 b; + Vec2 c; + Vec2 d; + + Quad() {} + Quad(const Vec2& a, const Vec2& b, const Vec2& c, const Vec2& d) + : a(a), b(b), c(c), d(d) {} + + void project(const Vec2& axis, float* min, float* max) const; + }; +} \ No newline at end of file diff --git a/public/blah/math/rect.cpp b/public/blah/math/rect.cpp new file mode 100644 index 0000000..1a2a293 --- /dev/null +++ b/public/blah/math/rect.cpp @@ -0,0 +1,111 @@ +#include +#include +#include +#include +#include +#include + +using namespace Blah; + +Rect::Rect() +{ + x = y = w = h = 0; +}; + +Rect::Rect(float rx, float ry, float rw, float rh) +{ + x = rx; + y = ry; + w = rw; + h = rh; +} + +Rect::Rect(Vec2 pos, Vec2 size) +{ + x = pos.x; + y = pos.y; + w = size.x; + h = size.y; +} + +Rect::Rect(RectI r) +{ + x = (float)r.x; + y = (float)r.y; + w = (float)r.w; + h = (float)r.h; +} + +Rect Rect::scale(float s) +{ + x = (x * s); + y = (y * s); + w = (w * s); + h = (h * s); + return *this; +} + +Rect Rect::scale(float sx, float sy) +{ + x = (x * sx); + y = (y * sy); + w = (w * sx); + h = (h * sy); + return *this; +} + +Line Rect::left_line() const { return Line(left(), top(), left(), bottom()); } +Line Rect::right_line() const { return Line(right(), top(), right(), bottom()); } +Line Rect::top_line() const { return Line(left(), top(), right(), top()); } +Line Rect::bottom_line() const { return Line(left(), bottom(), right(), bottom()); } + +Rect Rect::operator+(const Vec2& rhs) const { return Rect(x + rhs.x, y + rhs.y, w, h); } +Rect Rect::operator-(const Vec2& rhs) const { return Rect(x - rhs.x, y - rhs.y, w, h); } +Rect& Rect::operator+=(const Vec2& rhs) { x += rhs.x; y += rhs.y; return *this; } +Rect& Rect::operator-=(const Vec2& rhs) { x -= rhs.x; y -= rhs.y; return *this; } + +Rect Rect::transform(const Rect& rect, const Mat3x2& matrix) +{ + return Rect( + (rect.x * matrix.m11) + (rect.y * matrix.m21) + matrix.m31, + (rect.x * matrix.m12) + (rect.y * matrix.m22) + matrix.m32, + (rect.w * matrix.m11) + (rect.h * matrix.m21), + (rect.w * matrix.m12) + (rect.h * matrix.m22)); +} + +Rect Rect::transform(float x, float y, float w, float h, const Mat3x2& matrix) +{ + return Rect( + (x * matrix.m11) + (y * matrix.m21) + matrix.m31, + (x * matrix.m12) + (y * matrix.m22) + matrix.m32, + (w * matrix.m11) + (h * matrix.m21), + (w * matrix.m12) + (h * matrix.m22)); +} + +Rect Rect::from_points(Vec2& from, Vec2& to) +{ + Vec2 min = Vec2(Calc::min(from.x, to.x), Calc::min(from.y, to.y)); + Vec2 max = Vec2(Calc::max(from.x, to.x), Calc::max(from.y, to.y)); + return Rect(min.x, min.y, max.x - min.x, max.y - min.y); +} + +char Rect::get_sector(const Vec2& pt) const +{ + char h; + if (pt.x < left()) + h = 0b0001; + else if (pt.x >= right()) + h = 0b0010; + else + h = 0; + + char v; + if (pt.y < top()) + v = 0b0100; + else if (pt.y >= bottom()) + v = 0b1000; + else + v = 0; + + return h | v; +} \ No newline at end of file diff --git a/public/blah/math/rect.h b/public/blah/math/rect.h new file mode 100644 index 0000000..2c22a21 --- /dev/null +++ b/public/blah/math/rect.h @@ -0,0 +1,101 @@ +#pragma once +#include +#include +#include +#include + +namespace Blah +{ + struct Mat3x2; + + struct Rect + { + float x; + float y; + float w; + float h; + + Rect(); + Rect(float rx, float ry, float rw, float rh); + Rect(Vec2 pos, Vec2 size); + Rect(RectI r); + + Rect scale(float s); + Rect scale(float sx, float sy); + + float left() const { return x; } + float right() const { return x + w; } + float top() const { return y; } + float bottom() const { return y + h; } + + Vec2 center() const { return Vec2(x + w / 2, y + h / 2); } + float center_x() const { return x + w / 2; } + float center_y() const { return y + h / 2; } + + Vec2 top_left() const { return Vec2(x, y); } + Vec2 top_right() const { return Vec2(x + w, y); } + Vec2 bottom_right() const { return Vec2(x + w, y + h); } + Vec2 bottom_left() const { return Vec2(x, y + h); } + + Vec2 center_left() const { return Vec2(x, y + h / 2); } + Vec2 center_right() const { return Vec2(x + w, y + h / 2); } + Vec2 middle_top() const { return Vec2(x + w / 2, y); } + Vec2 middle_bottom() const { return Vec2(x + w / 2, y + h); } + + Line left_line() const; + Line right_line() const; + Line top_line() const; + Line bottom_line() const; + + bool contains(const Point& pt) const { return pt.x >= x && pt.x < x + w && pt.y >= y && pt.y < y + h; } + bool contains(const Vec2& pt) const { return pt.x >= x && pt.x < x + w && pt.y >= y && pt.y < y + h; } + + bool overlaps(const Rect& rect) const { return x + w >= rect.x && y + h >= rect.y && x < rect.x + rect.w && y < rect.y + rect.h; } + + bool intersects(const Line& line) const { return line.intersects(*this); } + bool intersects(const Line& line, Vec2* out_intersection_point) const { return line.intersects(*this, out_intersection_point); } + bool intersects(const Vec2& line_from, const Vec2& line_to) const { return intersects(Line(line_from, line_to)); } + bool intersects(const Vec2& line_from, const Vec2& line_to, Vec2* out_intersection_point) const { return intersects(Line(line_from, line_to), out_intersection_point); } + + Vec2 intersection_point(const Line& line) const + { + Vec2 ret; + if (line.intersects(*this, &ret)) + return ret; + else + return Vec2::zero; + } + + Vec2 intersection_point(const Vec2& line_from, const Vec2& line_to) const + { + Vec2 ret; + if (Line(line_from, line_to).intersects(*this, &ret)) + return ret; + else + return Vec2::zero; + } + + Rect inflate(float amount) { return Rect(x - amount, y - amount, w + amount * 2, h + amount * 2); } + + /* + Rect Sectors: + 0101 0100 0110 + 0001 0000 0010 + 1001 1000 1010 + 0000 = inside rectangle, all others refer to sectors relative to the rectangle + */ + char get_sector(const Vec2& pt) const; + + bool operator==(const Rect& rhs) const { return x == rhs.x && y == rhs.y && w == rhs.w && h == rhs.h; } + bool operator!=(const Rect& rhs) const { return !(*this == rhs); } + + Rect operator+(const Vec2& rhs) const; + Rect operator-(const Vec2& rhs) const; + Rect& operator+=(const Vec2& rhs); + Rect& operator-=(const Vec2& rhs); + + static Rect transform(const Rect& vec, const Mat3x2& matrix); + static Rect transform(float x, float y, float w, float h, const Mat3x2& matrix); + static Rect from_points(Vec2& from, Vec2& to); + }; +} diff --git a/public/blah/math/rectI.cpp b/public/blah/math/rectI.cpp new file mode 100644 index 0000000..9d01452 --- /dev/null +++ b/public/blah/math/rectI.cpp @@ -0,0 +1,90 @@ +#include +#include +#include +#include + +using namespace Blah; + +RectI::RectI() +{ + x = y = w = h = 0; +} + +RectI::RectI(int rx, int ry, int rw, int rh) +{ + x = rx; + y = ry; + w = rw; + h = rh; +} + +RectI::RectI(Point pos, Point size) +{ + x = pos.x; + y = pos.y; + w = size.x; + h = size.y; +} + +Point RectI::center() const { return Point(centerX(), centerY()); } +Point RectI::top_left() const { return Point(left(), top()); } +Point RectI::top_right() const { return Point(right(), top()); } +Point RectI::bottom_left() const { return Point(left(), bottom()); } +Point RectI::bottom_right() const { return Point(right(), bottom()); } + +bool RectI::contains(const Point& point) const +{ + return point.x >= x && point.x < x + w && point.y >= y && point.y < y + h; +} + +bool RectI::contains(const Vec2& point) const +{ + return point.x >= x && point.x < x + w && point.y >= y && point.y < y + h; +} + +char RectI::get_sector(const Point& pt) const +{ + char h; + if (pt.x < left()) + h = 0b0001; + else if (pt.x >= right()) + h = 0b0010; + else + h = 0; + + char v; + if (pt.y < top()) + v = 0b0100; + else if (pt.y >= bottom()) + v = 0b1000; + else + v = 0; + + return h | v; +} + +char RectI::get_sector(const Vec2& pt) const +{ + char h; + if (pt.x < left()) + h = 0b0001; + else if (pt.x >= right()) + h = 0b0010; + else + h = 0; + + char v; + if (pt.y < top()) + v = 0b0100; + else if (pt.y >= bottom()) + v = 0b1000; + else + v = 0; + + return h | v; +} + +RectI RectI::operator+(const Point& rhs) const { return RectI(x + rhs.x, y + rhs.y, w, h); } +RectI RectI::operator-(const Point& rhs) const { return RectI(x - rhs.x, y - rhs.y, w, h); } +RectI& RectI::operator+=(const Point& rhs) { x += rhs.x; y += rhs.y; return *this; } +RectI& RectI::operator-=(const Point& rhs) { x -= rhs.x; y -= rhs.y; return *this; } \ No newline at end of file diff --git a/public/blah/math/rectI.h b/public/blah/math/rectI.h new file mode 100644 index 0000000..2a8bfd5 --- /dev/null +++ b/public/blah/math/rectI.h @@ -0,0 +1,64 @@ +#pragma once + +namespace Blah +{ + struct Point; + struct Rect; + struct Vec2; + + struct RectI + { + int x; + int y; + int w; + int h; + + RectI(); + RectI(int rx, int ry, int rw, int rh); + RectI(Point pos, Point size); + + int left() const { return x; } + int right() const { return x + w; } + int top() const { return y; } + int bottom() const { return y + h; } + int centerX() const { return x + w / 2; } + int centerY() const { return y + h / 2; } + + Point center() const; + Point top_left() const; + Point top_right() const; + Point bottom_left() const; + Point bottom_right() const; + + bool overlaps(const RectI& other) const + { + return x < other.x + other.w + && other.x < x + w + && y < other.y + other.h + && other.y < y + h; + } + + bool contains(const Point& pt) const; + bool contains(const Vec2& pt) const; + + /* + Rect Sectors: + 0101 0100 0110 + 0001 0000 0010 + 1001 1000 1010 + 0000 = inside rectangle, all others refer to sectors relative to the rectangle + */ + char get_sector(const Point& pt) const; + char get_sector(const Vec2& pt) const; + + bool operator==(const RectI& rhs) const { return x == rhs.x && y == rhs.y && w == rhs.w && h == rhs.h; } + bool operator!=(const RectI& rhs) const { return !(*this == rhs); } + + RectI operator+(const Point& rhs) const; + RectI operator-(const Point& rhs) const; + RectI& operator+=(const Point& rhs); + RectI& operator-=(const Point& rhs); + + RectI operator*(const int& rhs) const { return RectI(x * rhs, y * rhs, w * rhs, h * rhs); } + }; +} \ No newline at end of file diff --git a/public/blah/math/stopwatch.cpp b/public/blah/math/stopwatch.cpp new file mode 100644 index 0000000..b28d39b --- /dev/null +++ b/public/blah/math/stopwatch.cpp @@ -0,0 +1,26 @@ +#include +#include + +using namespace std::chrono; +using namespace Blah; + +Stopwatch::Stopwatch() +{ + reset(); +} + +void Stopwatch::reset() +{ + start_time = std::chrono::duration_cast(system_clock::now().time_since_epoch()).count(); +} + +uint64_t Stopwatch::milliseconds() +{ + return microseconds() / 1000; +} + + +uint64_t Stopwatch::microseconds() +{ + return std::chrono::duration_cast(system_clock::now().time_since_epoch()).count() - start_time; +} diff --git a/public/blah/math/stopwatch.h b/public/blah/math/stopwatch.h new file mode 100644 index 0000000..1ea0b2a --- /dev/null +++ b/public/blah/math/stopwatch.h @@ -0,0 +1,16 @@ +#pragma once +#include + +namespace Blah +{ + class Stopwatch + { + public: + Stopwatch(); + void reset(); + uint64_t microseconds(); + uint64_t milliseconds(); + private: + uint64_t start_time; + }; +} \ No newline at end of file diff --git a/public/blah/math/vec2.cpp b/public/blah/math/vec2.cpp new file mode 100644 index 0000000..14718f0 --- /dev/null +++ b/public/blah/math/vec2.cpp @@ -0,0 +1,122 @@ +#include +#include +#include +#include + +using namespace Blah; + +Vec2::Vec2() { x = y = 0; } +Vec2::Vec2(float vx, float vy) { x = vx; y = vy; } +Vec2::Vec2(int vx, float vy) { x = (float)vx; y = vy; } +Vec2::Vec2(float vx, int vy) { x = vx; y = (float)vy; } +Vec2::Vec2(int vx, int vy) { x = (float)vx; y = (float)vy; } +Vec2::Vec2(Point p) { x = (float)p.x; y = (float)p.y; } + +Vec2 Vec2::operator +(const Vec2 rhs) const { return Vec2(x + rhs.x, y + rhs.y); } +Vec2 Vec2::operator -(const Vec2 rhs) const { return Vec2(x - rhs.x, y - rhs.y); } +Vec2 Vec2::operator /(const float rhs) const { return Vec2(x / rhs, y / rhs); } +Vec2 Vec2::operator *(const float rhs) const { return Vec2(x * rhs, y * rhs); } +Vec2 Vec2::operator-() const { return Vec2(-x, -y); } + +Vec2& Vec2::operator +=(const Vec2& rhs) { x += rhs.x; y += rhs.y; return *this; } +Vec2& Vec2::operator -=(const Vec2& rhs) { x -= rhs.x; y -= rhs.y; return *this; } +Vec2& Vec2::operator /=(const Vec2& rhs) { x /= rhs.x; y /= rhs.y; return *this; } +Vec2& Vec2::operator *=(const Vec2& rhs) { x *= rhs.x; y *= rhs.y; return *this; } +Vec2& Vec2::operator/=(float rhs) { x /= rhs; y /= rhs; return *this; } +Vec2& Vec2::operator*=(float rhs) { x *= rhs; y *= rhs; return *this; } + +bool Vec2::operator ==(const Vec2& rhs) { return x == rhs.x && y == rhs.y; } +bool Vec2::operator !=(const Vec2& rhs) { return x != rhs.x || y != rhs.y; } + +Vec2 Vec2::normal() const +{ + if (x == 0 && y == 0) + return zero; + float length = this->length(); + return Vec2(x / length, y / length); +} +float Vec2::length() const { return sqrtf(x * x + y * y); } +float Vec2::length_squared() const { return x * x + y * y; } + +Vec2 Vec2::perpendicular() const +{ + return Vec2(-y, x); +} + +float Vec2::angle() const +{ + return Calc::atan2(y, x); +} + +float Vec2::dot(Vec2 a, Vec2 b) { return (a.x * b.x + a.y * b.y); } +float Vec2::dot(float x, float y, Vec2 b) { return (x * b.x + y * b.y); } +float Vec2::dot(float x1, float y1, float x2, float y2) { return (x1 * x2 + y1 * y2); } + +Vec2 Vec2::transform(const Vec2& vec, const Mat3x2& matrix) +{ + return Vec2( + (vec.x * matrix.m11) + (vec.y * matrix.m21) + matrix.m31, + (vec.x * matrix.m12) + (vec.y * matrix.m22) + matrix.m32); +} + +Vec2 Vec2::transform(float x, float y, const Mat3x2& matrix) +{ + return Vec2( + (x * matrix.m11) + (y * matrix.m21) + matrix.m31, + (x * matrix.m12) + (y * matrix.m22) + matrix.m32); +} + +Vec2 Vec2::from_angle(float radians, float length) +{ + return Vec2((float)cos(radians) * length, (float)sin(radians) * length); +} + +Vec2 Vec2::from_angle(float radians) +{ + return from_angle(radians, 1); +} + +Vec2 Vec2::lerp(Vec2 a, Vec2 b, float t) +{ + if (t == 0) + return a; + else if (t == 1) + return b; + else + return a + (b - a) * t; +} + +Vec2 Vec2::bezier_lerp(Vec2 start, Vec2 b, Vec2 end, float t) +{ + return lerp(lerp(start, b, t), lerp(b, end, t), t); +} + +Vec2 Vec2::bezier_lerp(Vec2 start, Vec2 b, Vec2 c, Vec2 end, float t) +{ + return bezier_lerp(lerp(start, b, t), lerp(b, c, t), lerp(c, end, t), t); +} + +Vec2 Vec2::reflect(const Vec2& vector, const Vec2& normal) +{ + float dot = vector.x * normal.x + vector.y * normal.y; + + return Vec2( + vector.x - 2.0f * dot * normal.x, + vector.y - 2.0f * dot * normal.y); +} + +const Vec2 Vec2::unit_x = Vec2(1, 0); +const Vec2 Vec2::unit_y = Vec2(0, 1); +const Vec2 Vec2::right = Vec2(1, 0); +const Vec2 Vec2::up = Vec2(0, -1); +const Vec2 Vec2::down = Vec2(0, 1); +const Vec2 Vec2::left = Vec2(-1, 0); +const Vec2 Vec2::zero = Vec2(0, 0); +const Vec2 Vec2::one = Vec2(1, 1); + +#define DIAGONAL_UNIT 0.70710678118f +const Vec2 Vec2::up_right = Vec2(DIAGONAL_UNIT, -DIAGONAL_UNIT); +const Vec2 Vec2::up_left = Vec2(-DIAGONAL_UNIT, -DIAGONAL_UNIT); +const Vec2 Vec2::down_right = Vec2(DIAGONAL_UNIT, DIAGONAL_UNIT); +const Vec2 Vec2::down_left = Vec2(-DIAGONAL_UNIT, DIAGONAL_UNIT); +#undef DIAGONAL_UNIT \ No newline at end of file diff --git a/public/blah/math/vec2.h b/public/blah/math/vec2.h new file mode 100644 index 0000000..e505bbf --- /dev/null +++ b/public/blah/math/vec2.h @@ -0,0 +1,70 @@ +#pragma once +#include + +namespace Blah +{ + + struct Mat3x2; + + struct Vec2 + { + float x; + float y; + + Vec2(); + Vec2(float vx, float vy); + Vec2(int vx, float vy); + Vec2(float vx, int vy); + Vec2(int vx, int vy); + Vec2(Point p); + + Vec2 operator +(const Vec2 rhs) const; + Vec2 operator -(const Vec2 rhs) const; + Vec2 operator /(const float rhs) const; + Vec2 operator *(const float rhs) const; + Vec2 operator -() const; + + Vec2& operator +=(const Vec2& rhs); + Vec2& operator -=(const Vec2& rhs); + Vec2& operator /=(const Vec2& rhs); + Vec2& operator *=(const Vec2& rhs); + Vec2& operator /=(float rhs); + Vec2& operator *=(float rhs); + + bool operator ==(const Vec2& rhs); + bool operator !=(const Vec2& rhs); + + Vec2 normal() const; + float length() const; + float length_squared() const; + Vec2 perpendicular() const; + float angle() const; + + static float dot(Vec2 a, Vec2 b); + static float dot(float x, float y, Vec2 b); + static float dot(float x1, float y1, float x2, float y2); + + static Vec2 transform(const Vec2& vec, const Mat3x2& matrix); + static Vec2 transform(float x, float y, const Mat3x2& matrix); + static Vec2 from_angle(float radians, float length); + static Vec2 from_angle(float radians); + static Vec2 lerp(Vec2 start, Vec2 end, float t); + static Vec2 bezier_lerp(Vec2 start, Vec2 b, Vec2 end, float t); + static Vec2 bezier_lerp(Vec2 start, Vec2 b, Vec2 c, Vec2 end, float t); + static Vec2 reflect(const Vec2& vector, const Vec2& normal); + + static const Vec2 unit_x; + static const Vec2 unit_y; + static const Vec2 right; + static const Vec2 up; + static const Vec2 down; + static const Vec2 left; + static const Vec2 zero; + static const Vec2 one; + static const Vec2 up_right; + static const Vec2 up_left; + static const Vec2 down_right; + static const Vec2 down_left; + }; + +} \ No newline at end of file diff --git a/public/blah/math/vec4.h b/public/blah/math/vec4.h new file mode 100644 index 0000000..642c24b --- /dev/null +++ b/public/blah/math/vec4.h @@ -0,0 +1,15 @@ +#pragma once + +namespace Blah +{ + struct Vec4 + { + float x; + float y; + float z; + float w; + + Vec4() : x(0), y(0), z(0), w(0) {} + Vec4(float x, float y, float z, float w) : x(x), y(y), z(z), w(w) {} + }; +} \ No newline at end of file diff --git a/public/blah/streams/bufferstream.cpp b/public/blah/streams/bufferstream.cpp new file mode 100644 index 0000000..40ba0f0 --- /dev/null +++ b/public/blah/streams/bufferstream.cpp @@ -0,0 +1,107 @@ +#include +#include + +using namespace Blah; + +BufferStream::BufferStream() + : m_capacity(0), m_length(0), m_position(0) {} + +BufferStream::BufferStream(int capacity) + : m_capacity(0), m_length(0), m_position(0) +{ + if (capacity > 0) + { + m_buffer = new char[capacity]; + m_capacity = capacity; + } +} + +BufferStream::BufferStream(BufferStream&& src) noexcept +{ + m_buffer = src.m_buffer; + m_length = src.m_length; + m_capacity = src.m_capacity; + m_position = src.m_position; + src.m_buffer = nullptr; + src.m_position = src.m_length = src.m_capacity = 0; +} + +BufferStream& BufferStream::operator=(BufferStream&& src) noexcept +{ + m_buffer = src.m_buffer; + m_length = src.m_length; + m_capacity = src.m_capacity; + m_position = src.m_position; + src.m_buffer = nullptr; + src.m_position = src.m_length = src.m_capacity = 0; + return *this; +} + +BufferStream::~BufferStream() +{ + delete[] m_buffer; +} + +int64_t BufferStream::read_into(void* ptr, int64_t len) +{ + if (m_buffer == nullptr || ptr == nullptr) + return 0; + + if (len < 0) + return 0; + + if (len > m_length - m_position) + len = m_length - m_position; + + memcpy(ptr, m_buffer + m_position, (size_t)len); + m_position += len; + return len; +} + +int64_t BufferStream::write_from(const void* ptr, int64_t len) +{ + if (len < 0) + return 0; + + // resize + if (m_position + len >= m_capacity) + { + auto last_capacity = m_capacity; + + if (m_capacity <= 0) + m_capacity = 16; + while (m_position + len >= m_capacity) + m_capacity *= 2; + + char* new_buffer = new char[m_capacity]; + + if (m_buffer != nullptr) + { + memcpy(new_buffer, m_buffer, last_capacity); + delete[] m_buffer; + } + + m_buffer = new_buffer; + } + + // copy data + if (ptr != nullptr) + memcpy(m_buffer + m_position, ptr, (size_t)len); + + // increment position + m_position += len; + if (m_position > m_length) + m_length = m_position; + + // return the amount we wrote + return len; +} + +void BufferStream::close() +{ + delete[] m_buffer; + m_buffer = nullptr; + m_position = 0; + m_length = 0; + m_capacity = 0; +} \ No newline at end of file diff --git a/public/blah/streams/bufferstream.h b/public/blah/streams/bufferstream.h new file mode 100644 index 0000000..4a0303e --- /dev/null +++ b/public/blah/streams/bufferstream.h @@ -0,0 +1,37 @@ +#pragma once +#include + +namespace Blah +{ + class BufferStream : public Stream + { + public: + BufferStream(); + BufferStream(int capacity); + BufferStream(BufferStream&& bs) noexcept; + BufferStream& operator=(BufferStream&& bs) noexcept; + ~BufferStream(); + + virtual int64_t length() const override { return m_length; } + virtual int64_t position() const override { return m_position; } + virtual int64_t seek(int64_t seekTo) override { return m_position = (seekTo < 0 ? 0 : (seekTo > m_length ? m_length : seekTo)); } + virtual bool is_open() const override { return m_buffer != nullptr; } + virtual bool is_readable() const override { return true; } + virtual bool is_writable() const override { return true; } + virtual void close() override; + void clear() { m_length = m_position = 0; } + + char* data() { return m_buffer; } + const char* data() const { return m_buffer; } + + protected: + virtual int64_t read_into(void* ptr, int64_t length) override; + virtual int64_t write_from(const void* ptr, int64_t length) override; + + private: + char* m_buffer; + int64_t m_capacity; + int64_t m_length; + int64_t m_position; + }; +} \ No newline at end of file diff --git a/public/blah/streams/filestream.cpp b/public/blah/streams/filestream.cpp new file mode 100644 index 0000000..39b92fe --- /dev/null +++ b/public/blah/streams/filestream.cpp @@ -0,0 +1,96 @@ +#include +#include +#include + +using namespace Blah; + +FileStream::FileStream() +{ + m_handle = nullptr; + m_mode = FileMode::None; +} + +FileStream::FileStream(const char* path, FileMode mode) + : m_mode(mode) +{ + if (!Internal::Platform::file_open(path, &m_handle, mode)) + m_handle = nullptr; +} + +FileStream::FileStream(FileStream&& src) noexcept +{ + m_handle = src.m_handle; + m_mode = src.m_mode; + src.m_handle = nullptr; +} + +FileStream& FileStream::operator=(FileStream&& src) noexcept +{ + m_handle = src.m_handle; + m_mode = src.m_mode; + src.m_handle = nullptr; + return *this; +} + +FileStream::~FileStream() +{ + if (m_handle != nullptr) + Internal::Platform::file_close(m_handle); +} + +int64_t FileStream::length() const +{ + if (m_handle == nullptr) + return 0; + + return Internal::Platform::file_length(m_handle); +} + +int64_t FileStream::position() const +{ + if (m_handle == nullptr) + return 0; + + return Internal::Platform::file_position(m_handle); +} + +int64_t FileStream::seek(int64_t seek_to) +{ + if (m_handle == nullptr) + return 0; + + return Internal::Platform::file_seek(m_handle, seek_to); +} + +int64_t FileStream::read_into(void* ptr, int64_t length) +{ + if (m_handle == nullptr) + { + BLAH_ERROR("Unable to read from Stream"); + return 0; + } + + return Internal::Platform::file_read(m_handle, ptr, length); +} + +int64_t FileStream::write_from(const void* ptr, int64_t length) +{ + if (length <= 0) + return 0; + + if (m_handle == nullptr) + { + BLAH_ERROR("Unable to write to Stream"); + return 0; + } + + return Internal::Platform::file_write(m_handle, ptr, length); +} + +void FileStream::close() +{ + if (m_handle != nullptr) + Internal::Platform::file_close(m_handle); + m_handle = nullptr; + m_mode = FileMode::None; +} \ No newline at end of file diff --git a/public/blah/streams/filestream.h b/public/blah/streams/filestream.h new file mode 100644 index 0000000..85d63a5 --- /dev/null +++ b/public/blah/streams/filestream.h @@ -0,0 +1,32 @@ +#pragma once +#include +#include + +namespace Blah +{ + class FileStream : public Stream + { + public: + FileStream(); + FileStream(const char* path, FileMode mode = FileMode::ReadWrite); + FileStream(FileStream&& fs) noexcept; + FileStream& operator=(FileStream&& fs) noexcept; + ~FileStream(); + + virtual int64_t length() const override; + virtual int64_t position() const override; + virtual int64_t seek(int64_t seekTo) override; + virtual bool is_open() const override { return m_handle != nullptr; } + virtual bool is_readable() const override { return m_handle != nullptr && (m_mode == FileMode::ReadWrite || m_mode == FileMode::Read); } + virtual bool is_writable() const override { return m_handle != nullptr && (m_mode == FileMode::ReadWrite || m_mode == FileMode::Write); } + virtual void close() override; + + protected: + virtual int64_t read_into(void* ptr, int64_t length) override; + virtual int64_t write_from(const void* ptr, int64_t length) override; + + private: + FileMode m_mode; + void* m_handle; + }; +} \ No newline at end of file diff --git a/public/blah/streams/memorystream.cpp b/public/blah/streams/memorystream.cpp new file mode 100644 index 0000000..fafc807 --- /dev/null +++ b/public/blah/streams/memorystream.cpp @@ -0,0 +1,55 @@ +#include +#include + +using namespace Blah; + +MemoryStream::MemoryStream() + : m_data(nullptr), m_length(0), m_position(0) {} + +MemoryStream::MemoryStream(char* data, int64_t length) + : m_data(data), m_length(length), m_position(0) {} + +MemoryStream::MemoryStream(MemoryStream&& src) noexcept +{ + m_data = src.m_data; + m_position = src.m_position; + m_length = src.m_length; + src.m_data = nullptr; + src.m_length = src.m_position = 0; +} + +MemoryStream& MemoryStream::operator=(MemoryStream&& src) noexcept +{ + m_data = src.m_data; + m_position = src.m_position; + m_length = src.m_length; + src.m_data = nullptr; + src.m_length = src.m_position = 0; + return *this; +} + +int64_t MemoryStream::read_into(void* ptr, int64_t len) +{ + if (len < 0 || ptr == nullptr) + return 0; + + if (len > m_length - m_position) + len = m_length - m_position; + + memcpy(ptr, m_data + m_position, (size_t)len); + m_position += len; + return len; +} + +int64_t MemoryStream::write_from(const void* ptr, int64_t len) +{ + if (len < 0 || ptr == nullptr) + return 0; + + if (len > m_length - m_position) + len = m_length - m_position; + + memcpy(m_data + m_position, ptr, (size_t)len); + m_position += len; + return len; +} diff --git a/public/blah/streams/memorystream.h b/public/blah/streams/memorystream.h new file mode 100644 index 0000000..2fccf81 --- /dev/null +++ b/public/blah/streams/memorystream.h @@ -0,0 +1,35 @@ +#pragma once +#include + +namespace Blah +{ + class MemoryStream : public Stream + { + public: + MemoryStream(); + MemoryStream(char* data, int64_t length); + MemoryStream(MemoryStream&& ms) noexcept; + MemoryStream& operator=(MemoryStream&& ms) noexcept; + ~MemoryStream() { m_data = nullptr; m_length = m_position = 0; } + + virtual int64_t length() const override { return m_length; } + virtual int64_t position() const override { return m_position; } + virtual int64_t seek(int64_t seekTo) override { return m_position = (seekTo < 0 ? 0 : (seekTo > m_length ? m_length : seekTo)); } + virtual bool is_open() const override { return m_data != nullptr; } + virtual bool is_readable() const override { return true; } + virtual bool is_writable() const override { return true; } + virtual void close() override { m_data = nullptr; m_length = m_position = 0; } + + char* data() { return m_data; } + const char* data() const { return m_data; } + + protected: + virtual int64_t read_into(void* ptr, int64_t length) override; + virtual int64_t write_from(const void* ptr, int64_t length) override; + + private: + char* m_data; + int64_t m_length; + int64_t m_position; + }; +} \ No newline at end of file diff --git a/public/blah/streams/stream.cpp b/public/blah/streams/stream.cpp new file mode 100644 index 0000000..613475b --- /dev/null +++ b/public/blah/streams/stream.cpp @@ -0,0 +1,122 @@ +#include +#include +#include +#include + +using namespace Blah; + +Stream::Stream() +{ + +} + +int64_t Stream::pipe(Stream& stream, int64_t length) +{ + const int BUFFER_LENGTH = 4096; + int64_t result = 0; + + char buffer[BUFFER_LENGTH]; + while (length > 0) + { + auto step = length; + if (step > BUFFER_LENGTH) + step = BUFFER_LENGTH; + + auto count = read(buffer, step); + auto wrote = stream.write(buffer, count); + result += wrote; + length -= step; + + if (count < step || wrote < count) + break; + } + + return result; +} + +String Stream::read_line() +{ + String string; + read_line(string); + return string; +} + +int64_t Stream::read_line(String& writeTo) +{ + const int bufferSize = 512; + char buffer[bufferSize]; + + int64_t pos = position(); + int64_t length = 0; + int64_t count = 0; + bool hit = false; + + // read chunk-by-chunk + do + { + count = (int)read(buffer, bufferSize); + pos += count; + + // check for a newline + int64_t end = count; + for (int n = 0; n < count; n++) + if (buffer[n] == '\n' || buffer[n] == '\r') + { + hit = true; + end = n; + + // skip to the end of the line for future reading + int64_t lineEnd = pos - count + end + 1; + + // there might be a trailing '\n' + if (buffer[n] == '\r') + { + if (end < count && buffer[n + 1] == '\n') + { + lineEnd++; + } + // our buffer aligned perfectly ..... :/ + else if (count == bufferSize && end == count) + { + char ch; + if (read(&ch, 1) != 0 && ch == '\n') + lineEnd++; + } + } + + seek(lineEnd); + break; + } + + // copy to string + writeTo.set_length((int)(length + end)); + memcpy(writeTo.cstr() + length, buffer, (size_t)end); + *(writeTo.cstr() + length + end) = '\0'; + + // increment length + length += end; + + } while (!hit && count >= bufferSize); + + return length; +} + +String Stream::read_string(int length) +{ + if (length >= 0) + { + String str; + str.set_length(length); + read_into(str.cstr(), length); + str[length] = '\0'; + return str; + } + else + { + String str; + char next; + while (read_into(&next, 1) && next != '\0') + str.append(next); + return str; + } +} \ No newline at end of file diff --git a/public/blah/streams/stream.h b/public/blah/streams/stream.h new file mode 100644 index 0000000..13ea89b --- /dev/null +++ b/public/blah/streams/stream.h @@ -0,0 +1,108 @@ +#pragma once +#include +#include + +#define BLAH_SWAP_ENDIAN(value, type) \ + for (int i = 0; i < sizeof(type) / 2; i ++) { \ + uint8_t* _ptr = (uint8_t*)&value;\ + uint8_t _temp = *(_ptr + i); \ + *(_ptr + i) = *(_ptr + sizeof(type) - i - 1); \ + *(_ptr + sizeof(type) - i - 1) = _temp; \ + } + +#define BLAH_BIG_ENDIAN (*((short*)"AB") == 0x4243) + +namespace Blah +{ + enum class Endian + { + Little, + Big + }; + + class Stream + { + public: + Stream(); + Stream(const Stream&) = delete; + Stream& operator=(const Stream&) = delete; + + virtual ~Stream() = default; + + // returns the length of the stream + virtual int64_t length() const = 0; + + // returns the position of the stream + virtual int64_t position() const = 0; + + // seeks the position of the stream + virtual int64_t seek(int64_t seek_to) = 0; + + // returns true of the stream is open + virtual bool is_open() const = 0; + + // returns true of the stream is readable + virtual bool is_readable() const = 0; + + // returns true of the stream is writable + virtual bool is_writable() const = 0; + + // closes the stream + virtual void close() = 0; + + // pipes the contents of this tream to another stream + int64_t pipe(Stream& to, int64_t length); + + // reads a single line from this stream (up until \r or \n) + String read_line(); + + // reada a single line from this stream, to the given string (up until \r or \n) + int64_t read_line(String& writeTo); + + // reads a string of a given length, or until a null terminator if -1 + String read_string(int length = -1); + + // reads the amount of bytes into the given buffer, and returns the amount read + int64_t read(void* buffer, int64_t length) { return read_into(buffer, length); } + + template + T read() { return read(Endian::Little); } + + template + T read(Endian endian) + { + T value; + read(&value, sizeof(T)); + + if ((endian == Endian::Little && BLAH_BIG_ENDIAN) || (endian == Endian::Big && !BLAH_BIG_ENDIAN)) + BLAH_SWAP_ENDIAN(value, T); + + return value; + } + + int64_t write(const void* buffer, int64_t length) { return write_from(buffer, length); } + + template + int64_t write(const T& value) { return write(value, Endian::Little); } + + template + int64_t write(const T& value, Endian endian) + { + T writing = value; + if ((endian == Endian::Little && BLAH_BIG_ENDIAN) || (endian == Endian::Big && !BLAH_BIG_ENDIAN)) + BLAH_SWAP_ENDIAN(writing, T); + return write(&writing, sizeof(T)); + } + + protected: + // reads from the stream into the given buffer, and returns the number of bytes read + virtual int64_t read_into(void* buffer, int64_t length) = 0; + + // writes from the stream from the given buffer, and returns the number of bytes written + virtual int64_t write_from(const void* buffer, int64_t length) = 0; + + }; +} + +#undef BLAH_SWAP_ENDIAN +#undef BLAH_BIG_ENDIAN \ No newline at end of file diff --git a/public/blah/time.cpp b/public/blah/time.cpp new file mode 100644 index 0000000..f4bd8e4 --- /dev/null +++ b/public/blah/time.cpp @@ -0,0 +1,54 @@ +#include +#include + +using namespace Blah; + +namespace +{ + float modf(float x, float m) + { + return x - (int)(x / m) * m; + } +} + +uint64_t Time::milliseconds = 0; +float Time::elapsed = 0; +float Time::delta = 0; +float Time::previous_elapsed = 0; +float Time::pause_timer = 0; + +void Time::pause_for(float time) +{ + if (time >= pause_timer) + pause_timer = time; +} + +bool Time::on_interval(float time, float delta, float interval, float offset) +{ + return floor((time - offset - delta) / interval) < floor((time - offset) / interval); +} + +bool Time::on_interval(float delta, float interval, float offset) +{ + return Time::on_interval(Time::elapsed, delta, interval, offset); +} + +bool Time::on_interval(float interval, float offset) +{ + return Time::on_interval(Time::elapsed, Time::delta, interval, offset); +} + +bool Time::on_time(float time, float timestamp) +{ + return time >= timestamp && time - Time::delta < timestamp; +} + +bool Time::between_interval(float time, float interval, float offset) +{ + return modf(time - offset, interval * 2) >= interval; +} + +bool Time::between_interval(float interval, float offset) +{ + return between_interval(Time::elapsed, interval, offset); +} diff --git a/public/blah/time.h b/public/blah/time.h new file mode 100644 index 0000000..e4741e7 --- /dev/null +++ b/public/blah/time.h @@ -0,0 +1,34 @@ +#pragma once +#include + +namespace Blah +{ + struct Time + { + // uptime, in milliseconds + static uint64_t milliseconds; + + // uptime, in seconds + static float elapsed; + + // previous frame uptime, in seconds + static float previous_elapsed; + + // delta time from last frame + static float delta; + + // time the application should pause for + static float pause_timer; + + static void pause_for(float time); + + static bool on_interval(float time, float delta, float interval, float offset = 0); + static bool on_interval(float delta, float interval, float offset = 0); + static bool on_interval(float interval, float offset = 0); + + static bool on_time(float time, float timestamp); + + static bool between_interval(float time, float interval, float offset = 0); + static bool between_interval(float interval, float offset = 0); + }; +} \ No newline at end of file