fix(api): Handle invalid YAML input as HTTP 400

core/nurse_scheduling/loader.py

@@ -24,6 +24,11 @@
 
 yaml = YAML(typ='safe')
 
+
+class InputError(ValueError):
+    """Raised when user-provided input fails validation (HTTP 400 candidate)."""
+    pass
+
 def _load_yaml(content: bytes) -> Dict[str, Any]:
     """Load YAML from bytes content.
 
@@ -48,5 +53,17 @@ def load_data(content: bytes) -> NurseSchedulingData:
     Returns:
         NurseSchedulingData: The validated scheduling data
     """
-    data = _load_yaml(content)
-    return NurseSchedulingData(**data)
+    try:
+        data = _load_yaml(content)
+    except Exception as e:
+        raise InputError(f"Invalid YAML: {e}") from e
+    if not isinstance(data, dict):
+        raise InputError(
+            "Invalid YAML input: expected a mapping (key-value object) at the top level, "
+            f"but got {type(data).__name__}."
+        )
+    try:
+        return NurseSchedulingData(**data)
+    except TypeError as e:
+        # e.g. unexpected/missing keyword arguments from a malformed mapping
+        raise InputError(f"Invalid scheduling data: {e}") from e

core/nurse_scheduling/serve.py

@@ -25,6 +25,7 @@
 from fastapi.responses import StreamingResponse
 from fastapi.middleware.cors import CORSMiddleware
 from . import scheduler, exporter
+from .loader import InputError
 
 import sentry_sdk
 sentry_sdk.init(
@@ -146,6 +147,13 @@ async def optimize_and_export_xlsx(
             solver=solver,
         )
 
+    except InputError as e:
+        # Bad user input -> HTTP 400
+        logging.warning(f"Invalid input for optimization: {str(e)}")
+        raise HTTPException(
+            status_code=400,
+            detail=f"Invalid input: {str(e)}"
+        )
     except Exception as e:
         # TODO(security): Returning the error message to the client may be a security risk
         logging.error(f"Error during optimization: {str(e)}")