Core/Extractors: Fixed a possible issue with M2 model scaling

src/tools/vmap4_extractor/model.cpp

@@ -128,7 +128,6 @@ Vec3D fixCoordSystem2(Vec3D v)
 }
 
 ModelInstance::ModelInstance(MPQFile& f, char const* ModelInstName, uint32 mapID, uint32 tileX, uint32 tileY, FILE *pDirfile)
-    : model(NULL), d1(0), w(0.0f)
 {
     float ff[3];
     f.read(&id, 4);
@@ -136,7 +135,8 @@ ModelInstance::ModelInstance(MPQFile& f, char const* ModelInstName, uint32 mapID
     pos = fixCoords(Vec3D(ff[0], ff[1], ff[2]));
     f.read(ff, 12);
     rot = Vec3D(ff[0], ff[1], ff[2]);
-    f.read(&scale, 4);
+    f.read(&scale, 2);
+    f.read(&flags, 2);
     // scale factor - divide by 1024. blizzard devs must be on crack, why not just use a float?
     sc = scale / 1024.0f;