test: qualify immutable models and Phase 9 regression

docs/redesign/phase-9/p9-sec-02/full-model/prepare.py

@@ -0,0 +1,45 @@
+"""Prepare the reviewed trained checkpoint in a NEW nonproduction directory; never reset caches."""
+from pathlib import Path
+import hashlib
+import json
+import shutil
+import sys
+import time
+import urllib.parse
+import urllib.request
+
+sys.path.insert(0, str(Path(__file__).resolve().parent.parent/'model'))
+from verify import verify_fixture
+
+
+def prepare(destination):
+    destination=Path(destination)
+    manifest=json.loads(Path(__file__).with_name('ARTIFACTS.json').read_text())['weights-only']
+    revision=manifest['sourceRevision']
+    assert revision=='a4f8f3ea906ed274767e9906dbaede7531d660ff'
+    assert manifest['license'].startswith('Apache-2.0')
+    assert not destination.exists(), 'Use a new isolated directory; existing data is never replaced.'
+    assert shutil.disk_usage(destination.parent).free>8*1024**3, 'Require8GiB available for bounded preparation.'
+    destination.mkdir(mode=0o700)
+    deadline=time.monotonic()+1200
+    for entry in manifest['files']:
+        name=entry['file']; assert Path(name).name==name and name not in ('.','..')
+        url=entry.get('url') or entry['source']
+        parsed=urllib.parse.urlsplit(url)
+        assert parsed.scheme=='https' and parsed.netloc=='huggingface.co'
+        assert parsed.path.startswith('/sshleifer/distilbart-cnn-12-6/resolve/'+revision+'/')
+        temporary=destination/(name+'.partial'); count=0; digest=hashlib.sha256()
+        with urllib.request.urlopen(url,timeout=60) as response, temporary.open('xb') as stream:
+            while block:=response.read(1024*1024):
+                count+=len(block)
+                assert count<=entry['bytes'] and time.monotonic()<deadline, 'Artifact size/deadline exceeded.'
+                digest.update(block); stream.write(block)
+        assert count==entry['bytes'] and digest.hexdigest()==entry['sha256'], 'Artifact identity mismatch; do not load partial files.'
+        temporary.rename(destination/name)
+    verify_fixture(destination,manifest)
+    print(json.dumps({'status':'PASS','directory':str(destination.resolve()),'revision':revision,'files':len(manifest['files'])}))
+
+
+if __name__=='__main__':
+    assert len(sys.argv)==2, 'Usage: python prepare.py NEW_NONPRODUCTION_CACHE_DIRECTORY'
+    prepare(sys.argv[1])