I did some additional testing with example USDs exported from Houdini and the code works great! Thank you for this!
The code looks good in my initial review. I'll finish the review and testing tomorrow morning.
Since the tokens are declared static in this header file, I wonder if they should be inline static. I'm not sure if this is strictly necessary, but just a thought in terms of best practices.
This looks good to me as well. Thank you for this feature!
These changes look good on a high level and work well in the tests. I'll take a closer look at the details and hope to finish my review in the next day or two.
This looks good and is much cleaner now. Thank you!
I think there may be issues when exporting parented meshes. For example, if you export the attached mesh_hierarchy_instance.blend with instancing enabled to USDA, it looks like `/root/proto/Pla…
Okay, I think I was misunderstanding what you were doing. My comments were based on the design I originally had in mind, and you are doing something different. My apologies. Let me close my…
I deleted my last comment regarding the scene hierarchy, as I believe this was a local issue for me.
While testing, I've tried to use this with the example file provided at #126470 with the 2 instances in that file rotated (to give them transforms). When I Export with this PR in place, it seems…
Actually, we want to define the ref_path prim being referenced below, like this:
If you restore the original USDAbstractWriter::mark_as_instance() function you could replace this if clause with the following (though I haven't tested this):
It looks like you replaced this function with a comparable block of code in usd_writer_transform::do_write() but I wonder if this is necessary. For one thing, the new code introduces an error (see my comment below) but also the original function performs additional error checking.