Raise error if attachment pathname contains invalid bytes (python-discord#209)

Author: Esther-Goldberg

snekbox/nsjail.py

@@ -14,6 +14,7 @@
 from snekbox.limits.timed import time_limit
 from snekbox.result import EvalError, EvalResult
 from snekbox.snekio import FileAttachment, MemFS
+from snekbox.snekio.errors import IllegalPathError
 from snekbox.snekio.filesystem import Size
 from snekbox.utils.iter import iter_lstrip
 
@@ -244,6 +245,11 @@ def _parse_attachments(
         except TimeoutError as e:
             log.info(f"Exceeded time limit while parsing attachments: {e}")
             raise EvalError("TimeoutError: Exceeded time limit while parsing attachments") from e
+        except IllegalPathError as e:
+            log.info(f"Invalid bytes in filename while parsing attachments: {e}")
+            raise EvalError(
+                "FileParsingError: invalid bytes in filename while parsing attachments"
+            ) from e
         except Exception as e:
             log.exception(f"Unexpected {type(e).__name__} while parse attachments", exc_info=e)
             raise EvalError("FileParsingError: Unknown error while parsing attachments") from e

snekbox/snekio/attachment.py

@@ -67,8 +67,17 @@ def from_path(cls, file: Path, relative_to: Path | None = None) -> FileAttachmen
         Args:
             file: The file to attach.
             relative_to: The root for the path name.
+        Raises:
+            IllegalPathError: If path name contains characters that can't be encoded in UTF-8
         """
         path = file.relative_to(relative_to) if relative_to else file
+
+        # Disallow filenames with chars that can't be encoded in UTF-8
+        try:
+            str(path).encode("utf-8")
+        except UnicodeEncodeError as e:
+            raise IllegalPathError("File paths may not contain invalid byte sequences") from e
+
         return cls(str(path), file.read_bytes())
 
     @property

tests/test_nsjail.py

@@ -250,6 +250,20 @@ def test_file_parsing_timeout(self):
         )
         self.assertEqual(result.stderr, None)
 
+    def test_filename_encoding_illegal_chars(self):
+        code = dedent(
+            r"""
+            with open(b"\xC3.txt", "w") as f:
+                f.write("test")
+            """
+        ).strip()
+        result = self.eval_file(code)
+        self.assertEqual(result.returncode, None)
+        self.assertEqual(
+            result.stdout, "FileParsingError: invalid bytes in filename while parsing attachments"
+        )
+        self.assertEqual(result.stderr, None)
+
     def test_file_parsing_depth_limit(self):
         code = dedent(
             """