加载obj文件 DirectX VC++有注释

/* Demo Name: OBJ Model Loading Demo Author: Allen Sherrod Chapter: Ch 10 */ #include"objLoader.h" stObjModel *LoadOBJModel(char *fileName) { FILE *file; char *data = NULL; CToken lexer, tempLex; char tempLine[512]; char token[512]; // Open file for input. file = fopen(fileName, "r"); if(!file) return NULL; // Get the length of the file. fseek(file, 0, SEEK_END); int length = ftell(file); fseek(file, 0, SEEK_SET); // Read in all data from the file. data = new char[(length + 1) * sizeof(char)]; if(!data) return NULL; fread(data, length, 1, file); data[length] = '\0'; // Close the file when we are done. fclose(file); // Set our file to our lexer. lexer.SetTokenStream(data); // No longer need. delete[] data; data = NULL; bool validFile = false; // Look for the word Wavefront somewhere in the file to // determine if this .obj is compatiable since so modelers export // to slightly different formats. while(lexer.GetNextToken(token)) { if(strcmp(token, "Wavefront") == 0) { validFile = true; break; } } if(!validFile) return NULL; // Reset for next pass. lexer.Reset(); // Used to get the total number of each declared in a file. // Since faces uses indices these number could be different. int totalVertices = 0, totalNormals = 0, totalTexC = 0, totalFaces = 0; // Get the first (or next) line. while(lexer.MoveToNextLine(tempLine)) { // Set line to the temp lexer. tempLex.SetTokenStream(tempLine); // Read the new line character. lexer.GetNextToken(NULL); // If something was set to the temp lex then we keep going. if(!tempLex.GetNextToken(token)) continue; // If the first token of the line is a v, vn, vt, or f // increment the respective counter. if(strcmp(token, "v") == 0) totalVertices++; else if(strcmp(token, "vn") == 0) totalNormals++; else if(strcmp(token, "vt") == 0) totalTexC++; else if(strcmp(token, "f") == 0) totalFaces++; token[0] = '\0'; } // Allocate temp space to hold the data. Face are by 9 since there are // 3 vertices each with 3 values (v index/vt index/vn index). float *verts = new float[totalVertices * 3]; float *norms = new float[totalNormals * 3]; float *texC = new float[totalTexC * 2]; int *faces = new int[totalFaces * 9]; int vIndex = 0, nIndex = 0, tIndex = 0, fIndex = 0, index = 0; // Move to the beginning of the file. lexer.Reset(); // Do it all again but this time we get the data. while(lexer.MoveToNextLine(tempLine)) { // Set to temp lex, read past newline, get token. tempLex.SetTokenStream(tempLine); lexer.GetNextToken(NULL); if(!tempLex.GetNextToken(token)) continue; // If v then we get the vertex x, y, z. if(strcmp(token, "v") == 0) { // Get the x and save it. tempLex.GetNextToken(token); verts[vIndex] = (float)atof(token); vIndex++; // Get the y and save it. tempLex.GetNextToken(token); verts[vIndex] = (float)atof(token); vIndex++; // Get the z and save it. tempLex.GetNextToken(token); verts[vIndex] = (float)atof(token); vIndex++; } // Else If vn then we get the normal x, y, z. else if(strcmp(token, "vn") == 0) { // Get the x and save it. tempLex.GetNextToken(token); norms[nIndex] = (float)atof(token); nIndex++; // Get the y and save it. tempLex.GetNextToken(token); norms[nIndex] = (float)atof(token); nIndex++; // Get the z and save it. tempLex.GetNextToken(token); norms[nIndex] = (float)atof(token); nIndex++; } // Else If vt then we get the tex coord u, v. else if(strcmp(token, "vt") == 0) { // Get the u and save it. tempLex.GetNextToken(token); texC[tIndex] = (float)atof(token); tIndex++; // Get the v and save it. tempLex.GetNextToken(token); texC[tIndex] = (float)atof(token); tIndex++; } // Else If f then get each vertex 3 indices set. else if(strcmp(token, "f") == 0) { // Load for each vertex (3 in a triangle). for(int i = 0; i < 3; i++) { // Get first set. Get the length of it. tempLex.GetNextToken(token); int len = strlen(token); // Since there are no spaces between a set (1/1/1) // we can't simply read tokens so we must loop // through and take out each value before the / sign. for(int s = 0; s < len + 1; s++) { char buff[64]; // If this is not a / or if not at the end. if(token[s] != '/' && s < len) { buff[index] = token[s]; index++; } else { // Else end the string, convert it, save it. buff[index] = '\0'; faces[fIndex] = (int)atoi(buff); fIndex++; index = 0; } } } } token[0] = '\0'; } // No longer need. lexer.Shutdown(); // Create the model object by allocating. stObjModel *model = new stObjModel; if(!model) return NULL; memset(model, 0, sizeof(stObjModel)); // Save face count. model->numFaces = totalFaces; // Reset temp counters. vIndex = 0, nIndex = 0, tIndex = 0, fIndex = 0, index = 0; // Allocate data for each part of the model. model->vertices = new float[totalFaces * 3 * 3]; if(totalNormals) model->normals = new float[totalFaces * 3 * 3]; if(totalTexC) model->texCoords = new float[totalFaces * 3 * 2]; // Loop through and fill in our model. for(int f = 0; f < totalFaces * 9; f+=3) { // Get vertex. We subtract by 1 since we need our indices to be // in the range of 0 to max - 1 not 1 to max. We multiply by 3 // since there are 3 componets (x, y, z) in a vertex. model->vertices[vIndex + 0] = verts[(faces[f + 0] - 1) * 3 + 0]; model->vertices[vIndex + 1] = verts[(faces[f + 0] - 1) * 3 + 1]; model->vertices[vIndex + 2] = verts[(faces[f + 0] - 1) * 3 + 2]; vIndex += 3; // We do the same with the texture coordinate data. Since tex coord // data in the second thing we app 1 to f (v/vt/vn). if(model->texCoords) { model->texCoords[tIndex + 0] = texC[(faces[f + 1] - 1) * 2 + 0]; model->texCoords[tIndex + 1] = texC[(faces[f + 1] - 1) * 2 + 1]; tIndex += 2; } // We do the same with