create BMap project and copy-paste previous mesh convertion code

BMap/BMap.hpp

@@ -0,0 +1,129 @@
+#pragma once
+
+#include <VTUserAll.hpp>
+#include <vector>
+#include <cstdint>
+#include <cinttypes>
+
+namespace BMap {
+	
+	class BMGroup {
+	public:
+		BMGroup(LibCmo::CK2::ObjImpls::CKGroup* ptr);
+		~BMGroup();
+		LIBCMO_DEFAULT_COPY_MOVE(BMGroup);
+
+		LibCmo::CK2::ObjImpls::CKGroup* m_NativePtr;
+	};
+	
+	class BMTexture {
+	public:
+		BMTexture(LibCmo::CK2::ObjImpls::CKTexture* ptr);
+		~BMTexture();
+		LIBCMO_DEFAULT_COPY_MOVE(BMTexture);
+
+		LibCmo::CK2::ObjImpls::CKTexture* m_NativePtr;
+	};
+	
+	class BMMaterial {
+	public:
+		BMMaterial(LibCmo::CK2::ObjImpls::CKMaterial* ptr);
+		~BMMaterial();
+		LIBCMO_DEFAULT_COPY_MOVE(BMMaterial);
+
+		LibCmo::CK2::ObjImpls::CKMaterial* m_NativePtr;
+	};
+	
+	class BMMesh {
+	public:
+		BMMesh(LibCmo::CK2::ObjImpls::CKMesh* ptr);
+		~BMMesh();
+		LIBCMO_DEFAULT_COPY_MOVE(BMMesh);
+
+		LibCmo::CK2::ObjImpls::CKMesh* m_NativePtr;
+	};
+	
+	class BM3dEntity {
+	public:
+		BM3dEntity(LibCmo::CK2::ObjImpls::CK3dEntity* ptr);
+		~BM3dEntity();
+		LIBCMO_DEFAULT_COPY_MOVE(BM3dEntity);
+
+		LibCmo::CK2::ObjImpls::CK3dEntity* m_NativePtr;
+	};
+	
+	class BMMeshTransition {
+	private:
+		struct TransitionVertex {
+			TransitionVertex(
+				const LibCmo::VxMath::VxVector3& vec,
+				const LibCmo::VxMath::VxVector3& norm,
+				const LibCmo::VxMath::VxVector2& uv);
+			LibCmo::VxMath::VxVector3 m_Vertex;
+			LibCmo::VxMath::VxVector3 m_Norm;
+			LibCmo::VxMath::VxVector2 m_UV;
+		};
+		struct TransitionFace {
+			TransitionFace(uint32_t _i1, uint32_t _i2, uint32_t _i3, uint32_t mtl_id);
+			uint32_t m_Idx1, m_Idx2, m_Idx3;
+			uint32_t m_MtlSlotIdx;
+		};
+		struct TransitionVertexCompare {
+			bool operator()(const TransitionVertex& lhs, const TransitionVertex& rhs) const;
+		};
+
+	public:
+		BMMeshTransition();
+		~BMMeshTransition();
+
+		void PrepareVertexCount(uint32_t count);
+		void PrepareVertex(uint32_t index, float x, float y, float z);
+		void PrepareNormalCount(uint32_t count);
+		void PrepareNormal(uint32_t index, float x, float y, float z);
+		void PrepareUVCount(uint32_t count);
+		void PrepareUV(uint32_t index, float u, float v);
+		void PrepareMtlSlotCount(uint32_t count);
+		void PrepareMtlSlot(uint32_t index, BMMaterial* mtl);
+		void PrepareFaceCount(uint32_t count);
+		void PrepareFaceVertexIndices(uint32_t index, uint32_t indice1, uint32_t indice2, uint32_t indice3);
+		void PrepareFaceNormalIndices(uint32_t index, uint32_t indice1, uint32_t indice2, uint32_t indice3);
+		void PrepareFaceUVIndices(uint32_t index, uint32_t indice1, uint32_t indice2, uint32_t indice3);
+		void PrepareFaceMtlSlot(uint32_t index, uint32_t mtl_slot);
+
+		bool Parse(BMMesh* write_into_mesh);
+
+	private:
+		void DoRealParse();
+		void ApplyToMesh(BMMesh* write_into_mesh);
+
+		bool m_IsVertexOK, m_IsNormalOK, m_IsUVOK, m_IsFaceOK, m_IsMtlSlotOK;
+		bool m_IsParsed;
+
+		std::vector<LibCmo::VxMath::VxVector3> m_Vertexs, m_Normals;
+		std::vector<LibCmo::VxMath::VxVector2> m_UVs;
+		std::vector<uint32_t> m_FaceVertexs, m_FaceNormals, m_FaceUVs;
+		std::vector<uint32_t> m_FaceMtlSlotIdxs;
+		std::vector<BMMaterial*> m_MtlSlots;
+
+		std::vector<TransitionVertex> m_ProcVertexs;
+		std::vector<TransitionFace> m_ProcFaces;
+		// unordered_map have performance problem when dealing with massive data (in this case, big mesh)
+		// so we use map to get stable time cost.
+		std::map<TransitionVertex, uint32_t, TransitionVertexCompare> m_ProcDupRemover;
+	};
+
+	class BMFile {
+	public:
+		BMFile();
+		~BMFile();
+
+	private:
+		LibCmo::CK2::CKContext* m_Context;
+		std::vector<LibCmo::CK2::ObjImpls::CKGroup*> m_ObjGroups;
+		std::vector<LibCmo::CK2::ObjImpls::CK3dObject*> m_Obj3dObjects;
+		std::vector<LibCmo::CK2::ObjImpls::CKMesh*> m_ObjMeshs;
+		std::vector<LibCmo::CK2::ObjImpls::CKMaterial*> m_ObjMaterials;
+		std::vector<LibCmo::CK2::ObjImpls::CKTexture*> m_ObjTextures;
+	};
+
+}