feat: publish approved RF4 media catalog

rf4_research/media_assets.py

@@ -136,6 +136,27 @@ def reconcile_queued_duplicates(path: Path) -> dict:
     return {"duplicates": duplicates, "queued": sum(item.get("status") == "queued" for item in manifest.get("assets", []))}
 
 
+def approve_stored_assets(path: Path, *, note: str) -> dict:
+    """Publish every stored asset after an explicit owner-level approval."""
+    manifest = json.loads(path.read_text(encoding="utf-8"))
+    reviewed_at = datetime.now(timezone.utc).isoformat()
+    approved = 0
+    for item in manifest.get("assets", []):
+        if item.get("status") != "stored":
+            continue
+        stable_id = item.get("external_id") or item.get("sha256") or hashlib.sha256(item["asset_url"].encode()).hexdigest()
+        item.update({
+            "status": "approved",
+            "entity_key": f"{item.get('entity_type', 'reference')}:{stable_id}",
+            "reviewed_at": reviewed_at,
+            "review_note": note,
+        })
+        approved += 1
+    manifest["updated_at"] = reviewed_at
+    path.write_text(json.dumps(manifest, ensure_ascii=False, indent=2) + "\n", encoding="utf-8")
+    return {"approved": approved}
+
+
 def reclassify_manifest(path: Path) -> dict:
     manifest = json.loads(path.read_text(encoding="utf-8"))
     for item in manifest.get("assets", []):

rf4_research/media_cli.py

@@ -8,7 +8,7 @@ import urllib.error
 import urllib.request
 
 from .community_cli import MIN_FETCH_INTERVAL_SECONDS, USER_AGENT, _StrictRedirectHandler, _read_state, _validate_url_before_io, check_and_reserve, fetch_html, fetch_site_key
-from .media_assets import audit_media_catalog, extract_media_candidates, media_coverage, merge_manifest, reconcile_queued_duplicates, reclassify_manifest, review_asset, store_asset
+from .media_assets import approve_stored_assets, audit_media_catalog, extract_media_candidates, media_coverage, merge_manifest, reconcile_queued_duplicates, reclassify_manifest, review_asset, store_asset
 
 
 DEFAULT_ROOT = Path("data/media")
@@ -182,6 +182,7 @@ def main(argv: list[str] | None = None) -> int:
     parser.add_argument("--asset-url", help="Download this exact queued asset instead of the first queue item")
     parser.add_argument("--reclassify", action="store_true", help="Reapply current conservative classifier without network access")
     parser.add_argument("--dedupe-queue", action="store_true", help="Mark queued label matches to stored assets as duplicates without network access")
+    parser.add_argument("--approve-stored", action="store_true", help="Publish all stored assets after explicit owner approval")
     parser.add_argument("--review-url", help="Review an asset already present in the manifest")
     parser.add_argument("--decision", choices=("approved", "rejected"))
     parser.add_argument("--entity-type", choices=("fish", "waterbody", "tackle", "reference"))
@@ -217,6 +218,11 @@ def main(argv: list[str] | None = None) -> int:
         if args.dedupe_queue:
             print(json.dumps(reconcile_queued_duplicates(args.root / "manifest.json"), ensure_ascii=False, indent=2))
             return 0
+        if args.approve_stored:
+            if not args.note:
+                parser.error("--note is required with --approve-stored")
+            print(json.dumps(approve_stored_assets(args.root / "manifest.json", note=args.note), ensure_ascii=False, indent=2))
+            return 0
         if args.download_one:
             print(_download_one(args.root, args.state_file, args.asset_url))
             return 0