Remove macOS gcc compilation

I dont know why this keeps happening

.clang-format

@@ -1,6 +1,10 @@
 ---
-BasedOnStyle: LLVM
+BasedOnStyle: google
 IndentWidth: 4
 ColumnLimit: 160
+AllowShortBlocksOnASingleLine: Always
+AllowShortFunctionsOnASingleLine: All
+AlwaysBreakTemplateDeclarations: MultiLine
+#AlignConsecutiveDeclarations: true
 ---
 

.clang-tidy

@@ -1,21 +1,18 @@
----
-FormatStyle: LLVM
-Checks: 'clang-diagnostic-*,clang-analyzer-*,readability-*'
+FormatStyle: google
 
+Checks: '-*,clang-diagnostic-*,llvm-*,misc-*,-misc-unused-parameters,-misc-non-private-member-variables-in-classes,readability-identifier-naming,cppcoreguidelines-narrowing-conversions'
 CheckOptions:
-  - key: readability-identifier-naming.AbstractClassCase
-    value: 'CamelCase'
-  - key: readability-identifier-naming.ClassCase
-    value: 'CamelCase'
-  - key: readability-identifier-naming.EnumCase
-    value: 'CamelCase'
-  - key: readability-identifier-naming.StructCase
-    value: 'CamelCase'
-  - key: readability-identifier-naming.TypedefCase
-    value: 'CamelCase'
-  - key: readability-identifier-naming.MemberCase
-    value: 'CamelCase'
-  - key: readability-identifier-naming.FunctionCase
-    value: 'camelBack'
-  - key: readability-identifier-naming.ClassMethodCase
-    value: 'camelBack'
+  - { key: readability-identifier-naming.NamespaceCase,          value: lower_case }
+  - { key: readability-identifier-naming.ClassCase,              value: CamelCase }
+  - { key: readability-identifier-naming.StructCase,             value: CamelCase }
+  - { key: readability-identifier-naming.TemplateParameterCase,  value: CamelCase }
+  - { key: readability-identifier-naming.FunctionCase,           value: aNy_CasE }
+  - { key: readability-identifier-naming.VariableCase,           value: lower_case }
+  - { key: readability-identifier-naming.MemberCase,        value: lower_case }
+  - { key: readability-identifier-naming.PrivateMemberPrefix,    value: _ }
+  - { key: readability-identifier-naming.ProtectedMemberPrefix,  value: _ }
+  - { key: readability-identifier-naming.EnumConstantCase,         value: CamelCase }
+  - { key: readability-identifier-naming.ConstexprVariableCase,    value: CamelCase }
+  - { key: readability-identifier-naming.GlobalConstantCase,       value: CamelCase }
+  - { key: readability-identifier-naming.MemberConstantCase,       value: CamelCase }
+  - { key: readability-identifier-naming.StaticConstantCase,       value: CamelCase }

.github/workflows/cmake.yml

@@ -12,7 +12,26 @@ jobs:
     # well on Windows or Mac.  You can convert this to a matrix build if you need
     # cross-platform coverage.
     # See: https://docs.github.com/en/free-pro-team@latest/actions/learn-github-actions/managing-complex-workflows#using-a-build-matrix
-    runs-on: ubuntu-latest
+    runs-on: ${{ matrix.os }}
+    strategy:
+      matrix:
+        include:
+          - os: macos-latest
+            name: "macOS Clang"
+            cc: "clang"
+            cxx: "clang++"
+          - os: ubuntu-latest
+            name: "Ubuntu Clang"
+            cc: "clang-9"
+            cxx: "clang++-9"
+          - os: ubuntu-latest
+            name: "Ubuntu GCC"
+            cc: "gcc-10"
+            cxx: "g++-10"
+          - os: windows-latest
+            name: "Windows MSVC"
+            cc: "cl"
+            cxx: "cl"
 
     steps:
     - uses: actions/checkout@v2
@@ -21,6 +40,10 @@ jobs:
         # recursively checkout submodules.
         submodules: 'true'
 
+    - name: Set Windows ENV
+      if: runner.os == 'Windows'
+      uses: ilammy/msvc-dev-cmd@v1
+      
     - name: Create Build Environment
       # Some projects don't allow in-source building, so create a separate build directory
       # We'll use this as our working directory for all subsequent commands
@@ -34,11 +57,12 @@ jobs:
       # Note the current convention is to use the -S and -B options here to specify source 
       # and build directories, but this is only available with CMake 3.13 and higher.  
       # The CMake binaries on the Github Actions machines are (as of this writing) 3.12
-      run: cmake $GITHUB_WORKSPACE -DCMAKE_BUILD_TYPE=$BUILD_TYPE
+      run: cmake $GITHUB_WORKSPACE -DCMAKE_BUILD_TYPE=$BUILD_TYPE -DCMAKE_C_COMPILER=${{ matrix.cc }} -DCMAKE_CXX_COMPILER=${{ matrix.cxx }}
 
     - name: Build
       working-directory: ${{github.workspace}}/build
       shell: bash
+
       # Execute the build.  You can specify a specific target with "--target <NAME>"
       run: cmake --build . --config $BUILD_TYPE
 

.gitignore

@@ -17,4 +17,4 @@ install_manifest.txt
 compile_commands.json
 CTestTestfile.cmake
 _deps
-cmake-build-debug
+cmake-build-*

CMakeLists.txt

@@ -18,8 +18,7 @@ source_group(TREE ${CMAKE_CURRENT_SOURCE_DIR} FILES ${SOURCE_FILES} ${HEADER_FIL
 pybind11_add_module(python_rgbcx
         ${SOURCE_FILES}
         ${HEADER_FILES}
-        ${PYTHON_FILES}
-        )
+        ${PYTHON_FILES})
 
 add_executable(test_rgbcx
         ${SOURCE_FILES}
@@ -27,5 +26,11 @@ add_executable(test_rgbcx
         ${TEST_FILES})
 
 # Set module features, like C/C++ standards
-target_compile_features(python_rgbcx PUBLIC cxx_std_17 c_std_11)
-target_compile_features(test_rgbcx PUBLIC cxx_std_17 c_std_11)
+target_compile_features(python_rgbcx PUBLIC cxx_std_20 c_std_11)
+target_compile_features(test_rgbcx PUBLIC cxx_std_20 c_std_11)
+
+set_property(TARGET python_rgbcx test_rgbcx PROPERTY INTERPROCEDURAL_OPTIMIZATION True) #enable FLTO if available
+
+if (CMAKE_CXX_COMPILER_ID STREQUAL "Clang" AND CMAKE_SYSTEM_NAME STREQUAL "Darwin")
+    set_property(TARGET python_rgbcx test_rgbcx PROPERTY OSX_ARCHITECTURES_RELEASE x86_64 arm64) #Mach-O fat binary for arm and x86
+endif ()

src/blocks.cpp

@@ -18,38 +18,8 @@
  */
 
 #include "blocks.h"
+#include "color.h"
 #include <algorithm>
 #include <cassert>
 
-// region Color32 implementation
-Color32::Color32(uint32_t vr, uint32_t vg, uint32_t vb, uint32_t va) { set(vr, vg, vb, va); }
-
-uint8_t Color32::operator[](uint32_t idx) const {
-    assert(idx < 4);
-    return C[idx];
-}
-
-uint8_t &Color32::operator[](uint32_t idx) {
-    assert(idx < 4);
-    return C[idx];
-}
-
-void Color32::set(uint8_t vr, uint8_t vg, uint8_t vb, uint8_t va) {
-    C[0] = vr;
-    C[1] = vg;
-    C[2] = vb;
-    C[3] = va;
-}
-
-void Color32::set(const Color32 &other) {
-    C[0] = other.C[0];
-    C[1] = other.C[1];
-    C[2] = other.C[2];
-}
-Color32 Color32::min(const Color32 &a, const Color32 &b) {
-    return Color32(std::min(a[0], b[0]), std::min(a[1], b[1]), std::min(a[2], b[2]), std::min(a[3], b[3]));
-}
-Color32 Color32::max(const Color32 &a, const Color32 &b) {
-    return Color32(std::max(a[0], b[0]), std::max(a[1], b[1]), std::max(a[2], b[2]), std::max(a[3], b[3]));
-}
 // endregion

src/blocks.h

@@ -19,172 +19,120 @@
 
 #pragma once
 
-#include "util.h"
+#include <array>
 #include <cassert>
 #include <cstdint>
 #include <cstdlib>
 
-constexpr inline uint8_t DXT1SelectorBits = 2U;
+#include "color.h"
+#include "util.h"
 
 #pragma pack(push, 1)
-struct Color32 {
-    union {
-        struct {
-            uint8_t R;
-            uint8_t G;
-            uint8_t B;
-            uint8_t A;
-        };
+class BC1Block {
+   public:
+    uint16_t GetLowColor() const { return _low_color[0] | _low_color[1] << 8U; }
+    uint16_t GetHighColor() const { return _high_color[0] | _high_color[1] << 8U; }
+    Color32 GetLowColor32() const { return Color32::unpack565(GetLowColor()); }
+    Color32 GetHighColor32() const { return Color32::unpack565(GetHighColor()); }
 
-        uint8_t C[4];
-    };
-
-    Color32() {}
-
-    Color32(uint32_t vr, uint32_t vg, uint32_t vb, uint32_t va);
-
-    void set(uint8_t vr, uint8_t vg, uint8_t vb, uint8_t va);
-
-    void set(const Color32 &other);
-
-    uint8_t operator[](uint32_t idx) const;
-    uint8_t &operator[](uint32_t idx);
-
-    bool operator==(const Color32 &rhs) const {
-        return R == rhs.R && G == rhs.G && B == rhs.B && A == rhs.A;
+    bool Is3Color() const { return GetLowColor() <= GetHighColor(); }
+    void SetLowColor(uint16_t c) {
+        _low_color[0] = c & 0xFF;
+        _low_color[1] = (c >> 8) & 0xFF;
+    }
+    void SetHighColor(uint16_t c) {
+        _high_color[0] = c & 0xFF;
+        _high_color[1] = (c >> 8) & 0xFF;
+    }
+    uint32_t GetSelector(uint32_t x, uint32_t y) const {
+        assert((x < 4U) && (y < 4U));
+        return (selectors[y] >> (x * SelectorBits)) & SelectorMask;
+    }
+    void SetSelector(uint32_t x, uint32_t y, uint32_t val) {
+        assert((x < 4U) && (y < 4U) && (val < 4U));
+        selectors[y] &= (~(SelectorMask << (x * SelectorBits)));
+        selectors[y] |= (val << (x * SelectorBits));
     }
 
-    static Color32 min(const Color32 &a, const Color32 &b);
-    static Color32 max(const Color32 &a, const Color32 &b);
-};
-
-struct BC1Block {
     constexpr static inline size_t EndpointSize = 2;
     constexpr static inline size_t SelectorSize = 4;
     constexpr static inline uint8_t SelectorBits = 2;
     constexpr static inline uint8_t SelectorValues = 1 << SelectorBits;
     constexpr static inline uint8_t SelectorMask = SelectorValues - 1;
 
-    uint8_t LowColor[EndpointSize];
-    uint8_t HighColor[EndpointSize];
-    uint8_t Selectors[SelectorSize];
+   private:
+    std::array<uint8_t, EndpointSize> _low_color;
+    std::array<uint8_t, EndpointSize> _high_color;
 
-    inline uint32_t get_low_color() const { return LowColor[0] | (LowColor[1] << 8U); }
-    inline uint32_t get_high_color() const { return HighColor[0] | (HighColor[1] << 8U); }
-    inline bool is_3color() const { return get_low_color() <= get_high_color(); }
-    inline void set_low_color(uint16_t c) {
-        LowColor[0] = static_cast<uint8_t>(c & 0xFF);
-        LowColor[1] = static_cast<uint8_t>((c >> 8) & 0xFF);
-    }
-    inline void set_high_color(uint16_t c) {
-        HighColor[0] = static_cast<uint8_t>(c & 0xFF);
-        HighColor[1] = static_cast<uint8_t>((c >> 8) & 0xFF);
-    }
-    inline uint32_t get_selector(uint32_t x, uint32_t y) const {
-        assert((x < 4U) && (y < 4U));
-        return (Selectors[y] >> (x * SelectorBits)) & SelectorMask;
-    }
-    inline void set_selector(uint32_t x, uint32_t y, uint32_t val) {
-        assert((x < 4U) && (y < 4U) && (val < 4U));
-        Selectors[y] &= (~(SelectorMask << (x * SelectorBits)));
-        Selectors[y] |= (val << (x * DXT1SelectorBits));
-    }
-
-    static inline uint16_t pack_color(const Color32 &color, bool scaled, uint32_t bias = 127U) {
-        uint32_t r = color.R, g = color.G, b = color.B;
-        if (scaled) {
-            r = (r * 31U + bias) / 255U;
-            g = (g * 63U + bias) / 255U;
-            b = (b * 31U + bias) / 255U;
-        }
-        return static_cast<uint16_t>(minimum(b, 31U) | (minimum(g, 63U) << 5U) | (minimum(r, 31U) << 11U));
-    }
-
-    static inline uint16_t pack_unscaled_color(uint32_t r, uint32_t g, uint32_t b) { return static_cast<uint16_t>(b | (g << 5U) | (r << 11U)); }
-
-    static inline void unpack_color(uint32_t c, uint32_t &r, uint32_t &g, uint32_t &b) {
-        r = (c >> 11) & 31;
-        g = (c >> 5) & 63;
-        b = c & 31;
-
-        r = (r << 3) | (r >> 2);
-        g = (g << 2) | (g >> 4);
-        b = (b << 3) | (b >> 2);
-    }
-
-    static inline void unpack_color_unscaled(uint32_t c, uint32_t &r, uint32_t &g, uint32_t &b) {
-        r = (c >> 11) & 31;
-        g = (c >> 5) & 63;
-        b = c & 31;
-    }
+   public:
+    std::array<uint8_t, 4> selectors;
 };
 
-struct BC4Block {
+class BC4Block {
+   public:
+    inline uint32_t GetLowAlpha() const { return low_alpha; }
+    inline uint32_t GetHighAlpha() const { return high_alpha; }
+    inline bool Is6Alpha() const { return GetLowAlpha() <= GetHighAlpha(); }
+
+    inline uint64_t GetSelectorBits() const {
+        return ((uint64_t)((uint32_t)selectors[0] | ((uint32_t)selectors[1] << 8U) | ((uint32_t)selectors[2] << 16U) | ((uint32_t)selectors[3] << 24U))) |
+               (((uint64_t)selectors[4]) << 32U) | (((uint64_t)selectors[5]) << 40U);
+    }
+
+    inline uint32_t GetSelector(uint32_t x, uint32_t y, uint64_t selector_bits) const {
+        assert((x < 4U) && (y < 4U));
+        return (selector_bits >> (((y * 4) + x) * SelectorBits)) & (SelectorMask);
+    }
+
+    static inline std::array<uint8_t, 8> GetBlockValues6(uint32_t l, uint32_t h) {
+        return {static_cast<uint8_t>(l),
+                static_cast<uint8_t>(h),
+                static_cast<uint8_t>((l * 4 + h) / 5),
+                static_cast<uint8_t>((l * 3 + h * 2) / 5),
+                static_cast<uint8_t>((l * 2 + h * 3) / 5),
+                static_cast<uint8_t>((l + h * 4) / 5),
+                0,
+                255};
+    }
+
+    static inline std::array<uint8_t, 8> GetBlockValues8(uint32_t l, uint32_t h) {
+        return {static_cast<uint8_t>(l),
+                static_cast<uint8_t>(h),
+                static_cast<uint8_t>((l * 6 + h) / 7),
+                static_cast<uint8_t>((l * 5 + h * 2) / 7),
+                static_cast<uint8_t>((l * 4 + h * 3) / 7),
+                static_cast<uint8_t>((l * 3 + h * 4) / 7),
+                static_cast<uint8_t>((l * 2 + h * 5) / 7),
+                static_cast<uint8_t>((l + h * 6) / 7)};
+    }
+
+    static inline std::array<uint8_t, 8> GetBlockValues(uint32_t l, uint32_t h) {
+        if (l > h)
+            return GetBlockValues8(l, h);
+        else
+            return GetBlockValues6(l, h);
+    }
+
     constexpr static inline size_t EndpointSize = 1;
     constexpr static inline size_t SelectorSize = 6;
     constexpr static inline uint8_t SelectorBits = 3;
     constexpr static inline uint8_t SelectorValues = 1 << SelectorBits;
     constexpr static inline uint8_t SelectorMask = SelectorValues - 1;
 
-    uint8_t LowAlpha;
-    uint8_t HighAlpha;
-    uint8_t Selectors[SelectorSize];
-
-    inline uint32_t get_low_alpha() const { return LowAlpha; }
-    inline uint32_t get_high_alpha() const { return HighAlpha; }
-    inline bool is_alpha6_block() const { return get_low_alpha() <= get_high_alpha(); }
-
-    inline uint64_t get_selector_bits() const {
-        return ((uint64_t)((uint32_t)Selectors[0] | ((uint32_t)Selectors[1] << 8U) | ((uint32_t)Selectors[2] << 16U) |
-                           ((uint32_t)Selectors[3] << 24U))) |
-               (((uint64_t)Selectors[4]) << 32U) | (((uint64_t)Selectors[5]) << 40U);
-    }
-
-    inline uint32_t get_selector(uint32_t x, uint32_t y, uint64_t selector_bits) const {
-        assert((x < 4U) && (y < 4U));
-        return (selector_bits >> (((y * 4) + x) * SelectorBits)) & (SelectorMask);
-    }
-
-    static inline uint32_t get_block_values6(uint8_t *pDst, uint32_t l, uint32_t h) {
-        pDst[0] = static_cast<uint8_t>(l);
-        pDst[1] = static_cast<uint8_t>(h);
-        pDst[2] = static_cast<uint8_t>((l * 4 + h) / 5);
-        pDst[3] = static_cast<uint8_t>((l * 3 + h * 2) / 5);
-        pDst[4] = static_cast<uint8_t>((l * 2 + h * 3) / 5);
-        pDst[5] = static_cast<uint8_t>((l + h * 4) / 5);
-        pDst[6] = 0;
-        pDst[7] = 255;
-        return 6;
-    }
-
-    static inline uint32_t get_block_values8(uint8_t *pDst, uint32_t l, uint32_t h) {
-        pDst[0] = static_cast<uint8_t>(l);
-        pDst[1] = static_cast<uint8_t>(h);
-        pDst[2] = static_cast<uint8_t>((l * 6 + h) / 7);
-        pDst[3] = static_cast<uint8_t>((l * 5 + h * 2) / 7);
-        pDst[4] = static_cast<uint8_t>((l * 4 + h * 3) / 7);
-        pDst[5] = static_cast<uint8_t>((l * 3 + h * 4) / 7);
-        pDst[6] = static_cast<uint8_t>((l * 2 + h * 5) / 7);
-        pDst[7] = static_cast<uint8_t>((l + h * 6) / 7);
-        return 8;
-    }
-
-    static inline uint32_t get_block_values(uint8_t *pDst, uint32_t l, uint32_t h) {
-        if (l > h)
-            return get_block_values8(pDst, l, h);
-        else
-            return get_block_values6(pDst, l, h);
-    }
+    uint8_t low_alpha;
+    uint8_t high_alpha;
+    uint8_t selectors[SelectorSize];
 };
 
-struct BC3Block {
-    BC4Block AlphaBlock;
-    BC1Block ColorBlock;
+class BC3Block {
+   public:
+    BC4Block alpha_block;
+    BC1Block color_block;
 };
 
-struct BC5Block {
-    BC4Block RBlock;
-    BC4Block GBlock;
+class BC5Block {
+   public:
+    BC4Block r_block;
+    BC4Block g_block;
 };
-
 #pragma pack(pop)

src/color.cpp

@@ -0,0 +1,76 @@
+/*  Python-rgbcx Texture Compression Library
+    Copyright (C) 2021 Andrew Cassidy <drewcassidy@me.com>
+    Partially derived from rgbcx.h written by Richard Geldreich <richgel99@gmail.com>
+    and licenced under the public domain
+
+    This program is free software: you can redistribute it and/or modify
+    it under the terms of the GNU Affero General Public License as published by
+    the Free Software Foundation, either version 3 of the License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+    GNU Affero General Public License for more details.
+
+    You should have received a copy of the GNU Affero General Public License
+    along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include "color.h"
+#include <algorithm>
+#include <cassert>
+
+// region Color32 implementation
+Color32::Color32() { set(0, 0, 0, 0xFF); }
+
+Color32::Color32(uint8_t R, uint8_t G, uint8_t B, uint8_t A) { set(R, G, B, A); }
+
+uint16_t Color32::pack565Unscaled(uint16_t R, uint16_t G, uint16_t B) { return B | (G << 5) | (R << 11); }
+
+uint16_t Color32::pack565(uint16_t R, uint16_t G, uint16_t B) { return pack565Unscaled(scale8To5(R), scale8To6(G), scale8To5(B)); }
+
+Color32 Color32::unpack565(uint16_t Packed) {
+    uint8_t R = scale5To8((Packed >> 11) & 0x1F);
+    uint8_t G = scale6To8((Packed >> 5) & 0x3F);
+    uint8_t B = scale5To8(Packed & 0x1F);
+
+    return Color32(R, G, B);
+}
+
+uint8_t Color32::operator[](uint32_t Index) const {
+    assert(Index < 4);
+    return C[Index];
+}
+
+uint8_t &Color32::operator[](uint32_t Index) {
+    assert(Index < 4);
+    return C[Index];
+}
+
+void Color32::set(uint8_t R, uint8_t G, uint8_t B, uint8_t A) {
+    this->R = R;
+    this->G = G;
+    this->B = B;
+    this->A = A;
+}
+
+void Color32::set(const Color32 &Other) {
+    this->R = Other.R;
+    this->G = Other.G;
+    this->B = Other.B;
+    this->A = Other.A;
+}
+
+Color32 Color32::min(const Color32 &a, const Color32 &b) {
+    return Color32(std::min(a[0], b[0]), std::min(a[1], b[1]), std::min(a[2], b[2]), std::min(a[3], b[3]));
+}
+
+Color32 Color32::max(const Color32 &a, const Color32 &b) {
+    return Color32(std::max(a[0], b[0]), std::max(a[1], b[1]), std::max(a[2], b[2]), std::max(a[3], b[3]));
+}
+
+uint16_t Color32::pack565() { return pack565(R, G, B); }
+
+uint16_t Color32::pack565Unscaled() { return pack565Unscaled(R, G, B); }
+// endregion

src/color.h

@@ -0,0 +1,64 @@
+/*  Python-rgbcx Texture Compression Library
+    Copyright (C) 2021 Andrew Cassidy <drewcassidy@me.com>
+    Partially derived from rgbcx.h written by Richard Geldreich <richgel99@gmail.com>
+    and licenced under the public domain
+
+    This program is free software: you can redistribute it and/or modify
+    it under the terms of the GNU Affero General Public License as published by
+    the Free Software Foundation, either version 3 of the License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+    GNU Affero General Public License for more details.
+
+    You should have received a copy of the GNU Affero General Public License
+    along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#pragma once
+#include "util.h"
+#include <cassert>
+#include <cstdint>
+#include <cstdlib>
+
+#pragma pack(push, 1)
+class Color32 {
+  public:
+    union {
+        struct {
+            uint8_t R;
+            uint8_t G;
+            uint8_t B;
+            uint8_t A;
+        };
+
+        uint8_t C[4];
+    };
+
+    Color32();
+
+    Color32(uint8_t R, uint8_t G, uint8_t B, uint8_t A = 0xFF);
+
+    static uint16_t pack565Unscaled(uint16_t R, uint16_t G, uint16_t B);
+    static uint16_t pack565(uint16_t R, uint16_t G, uint16_t B);
+
+    static Color32 unpack565(uint16_t Packed);
+
+    bool operator==(const Color32 &Rhs) const { return R == Rhs.R && G == Rhs.G && B == Rhs.B && A == Rhs.A; }
+
+    uint8_t operator[](uint32_t Index) const;
+    uint8_t &operator[](uint32_t Index);
+
+    uint16_t pack565();
+    uint16_t pack565Unscaled();
+
+    static Color32 min(const Color32 &A, const Color32 &B);
+    static Color32 max(const Color32 &A, const Color32 &B);
+
+    void set(uint8_t R, uint8_t G, uint8_t B, uint8_t A);
+
+    void set(const Color32 &Other);
+};
+#pragma pack(pop)

src/rgbcx.h

@@ -55,6 +55,8 @@
 
 #include <cstdint>
 
+#include "blocks.h"
+
 // By default, the table used to accelerate cluster fit on 4 color blocks uses a 969x128 entry table.
 // To reduce the executable size, set RGBCX_USE_SMALLER_TABLES to 1, which selects the smaller 969x32 entry table.
 #ifndef RGBCX_USE_SMALLER_TABLES
@@ -173,8 +175,8 @@ void encode_bc1(void *pDst, const uint8_t *pPixels, uint32_t flags = 0, uint32_t
 // There are two encode_bc3() functions.
 // The first is the recommended function, which accepts a level parameter.
 // The second is a low-level version that allows fine control over BC1 encoding.
-void encode_bc3(uint32_t level, void *pDst, const uint8_t *pPixels);
-void encode_bc3(void *pDst, const uint8_t *pPixels, uint32_t flags = 0, uint32_t total_orderings_to_try = DEFAULT_TOTAL_ORDERINGS_TO_TRY);
+void encode_bc3(uint32_t level, BC3Block *pDst, const uint8_t *pPixels);
+void encode_bc3(BC3Block *pDst, const uint8_t *pPixels, uint32_t flags = 0, uint32_t total_orderings_to_try = DEFAULT_TOTAL_ORDERINGS_TO_TRY);
 
 // Encodes a single channel to BC4.
 // stride is the source pixel stride in bytes.
@@ -182,7 +184,7 @@ void encode_bc4(void *pDst, const uint8_t *pPixels, uint32_t stride = 4);
 
 // Encodes two channels to BC5.
 // chan0/chan1 control which channels, stride is the source pixel stride in bytes.
-void encode_bc5(void *pDst, const uint8_t *pPixels, uint32_t chan0 = 0, uint32_t chan1 = 1, uint32_t stride = 4);
+void encode_bc5(BC5Block *pDst, const uint8_t *pPixels, uint32_t chan0 = 0, uint32_t chan1 = 1, uint32_t stride = 4);
 
 // Decompression functions.
 
@@ -195,7 +197,7 @@ void unpack_bc4(const void *pBlock_bits, uint8_t *pPixels, uint32_t stride = 4);
 bool unpack_bc3(const void *pBlock_bits, void *pPixels, bc1_approx_mode mode = bc1_approx_mode::cBC1Ideal);
 
 void unpack_bc5(const void *pBlock_bits, void *pPixels, uint32_t chan0 = 0, uint32_t chan1 = 1, uint32_t stride = 4);
-} // namespace rgbcx
+}  // namespace rgbcx
 
 /*
 ------------------------------------------------------------------------------

src/test/test.cpp

@@ -20,10 +20,6 @@
 
 const int MAX_UBER_LEVEL = 5;
 
-inline int iabs(int i) { if (i < 0) i = -i; return i; }
-inline uint8_t clamp255(int32_t i) { return (uint8_t)((i & 0xFFFFFF00U) ? (~(i >> 31)) : i); }
-template <typename S> inline S clamp(S value, S low, S high) { return (value < low) ? low : ((value > high) ? high : value); }
-
 static int print_usage()
 {
 	fprintf(stderr, "bc7enc\n");
@@ -781,7 +777,7 @@ int main(int argc, char *argv[])
 			}
 			case DXGI_FORMAT_BC3_UNORM:
 			{
-				block16* pBlock = &packed_image16[bx + by * blocks_x];
+				BC3Block* pBlock = reinterpret_cast<BC3Block *>(&packed_image16[bx + by * blocks_x]);
 
 				rgbcx::encode_bc3(bc1_quality_level, pBlock, &pixels[0].m_c[0]);
 				break;
@@ -797,7 +793,7 @@ int main(int argc, char *argv[])
 			{
 				block16* pBlock = &packed_image16[bx + by * blocks_x];
 
-				rgbcx::encode_bc5(pBlock, &pixels[0].m_c[0], bc45_channel0, bc45_channel1, 4);
+				rgbcx::encode_bc5(reinterpret_cast<BC5Block *>(pBlock), &pixels[0].m_c[0], bc45_channel0, bc45_channel1, 4);
 				break;
 			}
 			case DXGI_FORMAT_BC7_UNORM:

src/util.h

@@ -23,17 +23,17 @@
 static inline uint32_t iabs(int32_t i) { return (i < 0) ? static_cast<uint32_t>(-i) : static_cast<uint32_t>(i); }
 static inline uint64_t iabs(int64_t i) { return (i < 0) ? static_cast<uint64_t>(-i) : static_cast<uint64_t>(i); }
 
-static inline uint8_t scale_8_to_5(uint32_t v) {
+static inline uint8_t scale8To5(uint32_t v) {
     v = v * 31 + 128;
     return (uint8_t)((v + (v >> 8)) >> 8);
 }
-static inline uint8_t scale_8_to_6(uint32_t v) {
+static inline uint8_t scale8To6(uint32_t v) {
     v = v * 63 + 128;
     return (uint8_t)((v + (v >> 8)) >> 8);
 }
 
-static inline int scale_5_to_8(int v) { return (v << 3) | (v >> 2); }
-static inline int scale_6_to_8(int v) { return (v << 2) | (v >> 4); }
+static inline int scale5To8(int v) { return (v << 3) | (v >> 2); }
+static inline int scale6To8(int v) { return (v << 2) | (v >> 4); }
 
 template <typename S> inline S maximum(S a, S b) { return (a > b) ? a : b; }
 template <typename S> inline S maximum(S a, S b, S c) { return maximum(maximum(a, b), c); }
@@ -55,7 +55,7 @@ static inline float clampf(float value, float low, float high) {
 static inline uint8_t clamp255(int32_t i) { return (uint8_t)((i & 0xFFFFFF00U) ? (~(i >> 31)) : i); }
 
 template <typename S> inline S clamp(S value, S low, S high) { return (value < low) ? low : ((value > high) ? high : value); }
-static inline int32_t clampi(int32_t value, int32_t low, int32_t high) {
+static inline int32_t          clampi(int32_t value, int32_t low, int32_t high) {
     if (value < low)
         value = low;
     else if (value > high)
@@ -67,8 +67,3 @@ static inline int squarei(int a) { return a * a; }
 static inline int absi(int a) { return (a < 0) ? -a : a; }
 
 template <typename F> inline F lerp(F a, F b, F s) { return a + (b - a) * s; }
-
-
-
-
-