feat: track media catalog coverage

rf4_research/media_cli.py

@@ -8,7 +8,7 @@ import urllib.error
 import urllib.request
 
 from .community_cli import USER_AGENT, _StrictRedirectHandler, _validate_url_before_io, check_and_reserve, fetch_html, fetch_site_key
-from .media_assets import audit_media_catalog, extract_media_candidates, merge_manifest, reclassify_manifest, review_asset, store_asset
+from .media_assets import audit_media_catalog, extract_media_candidates, media_coverage, merge_manifest, reclassify_manifest, review_asset, store_asset
 
 
 DEFAULT_ROOT = Path("data/media")
@@ -59,6 +59,7 @@ def main(argv: list[str] | None = None) -> int:
     parser.add_argument("--entity-key")
     parser.add_argument("--note")
     parser.add_argument("--audit", action="store_true", help="Verify manifest metadata, hashes and local files without network access")
+    parser.add_argument("--coverage", action="store_true", help="Compare candidates and approvals with the catalog baseline")
     parser.add_argument("--root", type=Path, default=DEFAULT_ROOT)
     parser.add_argument("--state-file", type=Path, default=Path(".cache/community-fetch-state.json"))
     args = parser.parse_args(argv)
@@ -67,6 +68,9 @@ def main(argv: list[str] | None = None) -> int:
             report = audit_media_catalog(args.root)
             print(json.dumps(report, ensure_ascii=False, indent=2))
             return 1 if report["issues"] or report["orphaned_files"] else 0
+        if args.coverage:
+            print(json.dumps(media_coverage(args.root), ensure_ascii=False, indent=2))
+            return 0
         if args.review_url:
             if not args.decision:
                 parser.error("--decision is required with --review-url")