scripts/prepare_samples.py

"""Download a small, real validation set. Not the catalogue limit."""
import json, urllib.request, concurrent.futures
from pathlib import Path
ROOT = Path(__file__).resolve().parents[1]
manifest_path=ROOT/'.cache/media-manifest.json'
if manifest_path.exists():manifest=json.loads(manifest_path.read_text())
else:
    manifest=json.load(urllib.request.urlopen('https://nvidia-simready-central.hf.space/library_media/usdsearch-dev/manifest.json',timeout=60))
    manifest_path.parent.mkdir(exist_ok=True);manifest_path.write_text(json.dumps(manifest))
terms = ['food_produce_banana_m01/', 'chair_office_fabric_manager_a01/', 'Robotiq_2F_140/simready_usd/Robotiq_2F_140.usda', 'dishwasher_ldfn4542b/', 'cooking_ldgl6924s/', 'chest_of_drawers/']
samples = []
for term in terms:
    source, asset = next((k, v) for k, v in manifest['assets'].items() if term in k and v.get('usd'))
    base = 'https://nvidia-simready-central.hf.space/library_media/usdsearch-dev/' + asset['id'] + '/'
    name = source.split('/blob/')[1].split('/')[1]
    samples.append(dict(id=asset['id'], name=name, source=source, url=base+asset['usd']['files']['asset'], meta=base+asset['usd']['files']['articulation'], compact=asset['usd'].get('compact'), moving=asset.get('moving', 0)))
(ROOT/'public/samples.json').write_text(json.dumps(samples, indent=2))
def get(sample):
    folder = ROOT / '.cache/samples' / sample['id']; folder.mkdir(parents=True, exist_ok=True)
    for url, name in [(sample['url'], 'asset.usdc'), (sample['meta'], 'articulation.json')]:
        dest = folder / name
        if not dest.exists():
            dest.write_bytes(urllib.request.urlopen(url, timeout=60).read())
    print(sample['name'], (folder/'asset.usdc').stat().st_size, flush=True)
with concurrent.futures.ThreadPoolExecutor(max_workers=4) as pool:
    list(pool.map(get, samples))