Added more encoding options

submit.py

@@ -57,6 +57,16 @@ def upload_submission(
     Path(tmp_name).unlink()
 
 
+def safe_read_encoding(path_obj: Path) -> str:
+    # Try utf-8 otherwise latin-1 (can tweak as needed)
+    try:
+        with path_obj.open("rb") as f:
+            return f.read().decode("utf-8")
+    except UnicodeDecodeError:
+        with path_obj.open("rb") as f:
+            return f.read().decode("latin-1")
+
+
 def validate_file_requirements(file: BinaryIO, file_type: str) -> Path:
     """Validate basic file requirements and return Path object"""
     file_path = file.name
@@ -70,6 +80,13 @@ def validate_file_requirements(file: BinaryIO, file_type: str) -> Path:
         raise gr.Error(
             f"{file_type} file must be a CSV file. Please upload a .csv file."
         )
+    # Check it can be read
+    try:
+        safe_read_encoding(path_obj)
+    except Exception as e:
+        raise gr.Error(
+            f"Error reading {file_type} file {path_obj.name}. Could be an encoding issue if using Windows. Full error: {e}"
+        )
 
     return path_obj
 
@@ -115,15 +132,13 @@ def make_submission(
     files = {}
     # Validate CV file
     cv_path = validate_file_requirements(cv_file, "GDPa1 Cross-Validation")
-    with cv_path.open("rb") as f:
-        cv_content = f.read().decode("utf-8")
+    cv_content = safe_read_encoding(cv_path)
     validate_csv_file(cv_content, "GDPa1_cross_validation")
     files["cv"] = cv_content
 
     # Validate test file
     test_path = validate_file_requirements(test_file, "Private Test Set")
-    with test_path.open("rb") as f:
-        test_content = f.read().decode("utf-8")
+    test_content = safe_read_encoding(test_path)
     validate_csv_file(test_content, "Heldout Test Set")
     files["test"] = test_content