将BMAX LOD逻辑应用到C++

[LiXizhi]

放到这里， 默认开启。

注意不需要转ParaX文件了， 直接变成1-3个内存对象即可。

[LiXizhi]

BMAXParser 接口

BMaxParser p(myFile.getBuffer(), myFile.getSize());
-- calling the ParseParaXModel the first time does not process LOD
lod.m_pParaXMesh = p.ParseParaXModel();
if( lod.m_pParaXMesh.GetVerticeCount() > XXXX)  then
     -- add a new lod : only process LOD related info, when it is queried the first time. 
    lod1.m_pParaXMesh = p.ParseLODParaXModel(nMaxTriangleCount);
     --- 
    lod2.m_pParaXMesh = p.ParseLODParaXModel(nMaxTriangleCount);
end

[LiXizhi]

// block max model. 
BMaxParser p(myFile.getBuffer(), myFile.getSize());
lod.m_pParaXMesh = p.ParseParaXModel();
auto pParaXMesh = lod.m_pParaXMesh;
if (m_MeshLODs.size() == 1)
{
	// each LOD at least cut triangle count in half and no bigger than a given count. 
	const int nLodsMaxTriangleCounts[] = { 2000, 500, 100};
	
	for (int i = 0; pParaXMesh && i < sizeof(nLodsMaxTriangleCounts)/sizeof(int); i++)
	{
		if ((int)pParaXMesh->GetPolyCount() >= nLodsMaxTriangleCounts[i])
		{
			MeshLOD lod;
			lod.m_pParaXMesh = p.ParseParaXModel(std::min(nLodsMaxTriangleCounts[i], (int)(pParaXMesh->GetObjectNum().nVertices / 2)));
			lod.m_fromDepthSquared = (float)(((30+i)*30) ^ 2);
			if (lod.m_pParaXMesh)
			{
				pParaXMesh = lod.m_pParaXMesh;
				pMeshLODs.push_back(lod);
			}
		}
	}
}

This is how i used it. but...

遇到一个问题。 是否可以反复的使用p.ParserParaXModel(n)
可能前一次的中间数据结构没有清空。 导致，每次调用Triangle数量反而增多了。

@Winless