DAEParser not parsing visual_scenes or controllers / morph correctly

[bakajin]

Of course this parser is new, I just have a few things that would be good if they made it there. I guess a feature request, as I do have workarounds for this.

So far I can see there are some things missing relating to the parsing of visual_scenes. Mainly the actual names relating to the library_geometry ids (this helps with hooking up animation and my next point).

IE when the DAE states and in that node states <instance_geometry url="#geometries_00"> or <instance_controller url="#controllers_0">.

The parsing of the library_controllers, mainly the controller morph.

In my DAEs the library_geometries contains geometries with only a list of vertices in a float_array, which should be enough to set up vertexAnimation. The name of the base model and the list of these geometry names is stored in library_controllers

Like:

geometries_49 geometries_50 geometries_51 geometries_52 geometries_53 geometries_54 geometries_55 geometries_56 geometries_57 geometries_58 geometries_59 geometries_60 geometries_61 geometries_62 geometries_63

etc.

However I am a bit in the dark as to how the animation/timing of the vertexAnimation/morph/shape animation is actually stored in the DAE. It seems to me there is actually no way of writing this to into a DAE. But the keyframes are a start, one could automagically name them the framenumber in the 3d software I guess. I so desired I'll gladly send some DAEs with these data for testing.

Also it would be great if the scene animation could somehow be made available by the parser, mainly lists of floats with xyz and time info. Now it seems I would have to parse the DAE file seperatly as XML to retrieve that data. Maybe an arbritrary library where certain data can be retrieved? Animation also uses the model names from the visual_scenes as reference btw.

But great work, I have my workarounds and might hazzard a go to add the previous functionality myself. Just thought it prudent to point it out.

Thanks

O

[richardolsson]

@Fabrice3D Assigning this to you. Lets try to get any relevant improvements in for 4.1 if we can? Do you agree?