Merge nucleic/sleek-ember-seal-uady into dev

review_data.py

@@ -65,6 +65,17 @@ def reviewed_population(args: argparse.Namespace) -> list[SourceRecord]:
     ).records
 
 
+def review_policy_status(path: Path) -> str:
+    try:
+        value = json.loads(path.read_text(encoding="utf-8"))
+        status = value["humanLabelAndDifficultyReview"]["status"]
+    except (OSError, UnicodeError, json.JSONDecodeError, KeyError, TypeError) as exc:
+        raise DataError(f"{path}: cannot read human-review policy: {exc}") from exc
+    if not isinstance(status, str) or not status:
+        raise DataError(f"{path}: invalid human-review policy status")
+    return status
+
+
 def build_parser() -> argparse.ArgumentParser:
     parser = argparse.ArgumentParser(description=__doc__)
     parser.add_argument("--source", action="append", type=Path)
@@ -98,6 +109,7 @@ def main(argv: Sequence[str] | None = None) -> int:
     args = build_parser().parse_args(argv)
     try:
         population = reviewed_population(args)
+        policy_status = review_policy_status(args.curation_review.resolve())
         sample = stratified_review_sample(
             population,
             fraction=args.review_fraction,
@@ -152,7 +164,8 @@ def main(argv: Sequence[str] | None = None) -> int:
     print(
         f"Human review: {progress.completed}/{progress.records} complete "
         f"({progress.accepted} accept, {progress.relabeled} relabel, "
-        f"{progress.rejected} reject, {progress.incomplete} remaining)."
+        f"{progress.rejected} reject, {progress.incomplete} remaining). "
+        f"Dataset policy: {policy_status}."
     )
     return 0