FIX: Fix DBNStore exception on empty data

Author: nmacholl

CHANGELOG.md

@@ -2,6 +2,7 @@
 
 ## 0.14.1 - TBD
 - Fixed issue where `DBNStore.to_df()` would raise an exception if no records were present
+- Fixed exception message when creating a DBNStore from an empty data source
 
 ## 0.14.0 - 2023-06-14
 - Added support for reusing a `Live` client to reconnect

databento/common/dbnstore.py

@@ -139,6 +139,11 @@ def __init__(self, source: PathLike[str] | str):
         if not self._path.is_file() or not self._path.exists():
             raise FileNotFoundError(source)
 
+        if self._path.stat().st_size == 0:
+            raise ValueError(
+                f"Cannot create data source from empty file: {self._path.name}",
+            )
+
         self._name = self._path.name
         self.__buffer: IO[bytes] | None = None
 
@@ -308,6 +313,11 @@ class DBNStore:
     to_ndarray : np.ndarray
         The data as a numpy `ndarray`.
 
+    Raises
+    ------
+    BentoError
+        When the data_source does not contain valid DBN data or is corrupted.
+
     See Also
     --------
     https://docs.databento.com/knowledge-base/new-users/dbn-encoding
@@ -330,7 +340,7 @@ def __init__(self, data_source: DataSource) -> None:
             buffer = data_source.reader
         else:
             # We don't know how to read this file
-            raise RuntimeError(
+            raise BentoError(
                 f"Could not determine compression format of {self._data_source.name}",
             )
 
@@ -736,7 +746,9 @@ def from_file(cls, path: PathLike[str] | str) -> DBNStore:
         Raises
         ------
         FileNotFoundError
-            If a empty or non-existant file is specified.
+            If a non-existant file is specified.
+        ValueError
+            If an empty file is specified.
 
         """
         return cls(FileDataSource(path))
@@ -757,8 +769,8 @@ def from_bytes(cls, data: BytesIO | bytes | IO[bytes]) -> DBNStore:
 
         Raises
         ------
-        FileNotFoundError
-            If a empty or non-existant file is specified.
+        ValueError
+            If an empty buffer is specified.
 
         """
         return cls(MemoryDataSource(data))

tests/test_historical_bento.py

@@ -3,6 +3,7 @@
 import collections
 import datetime as dt
 import sys
+from io import BytesIO
 from pathlib import Path
 from typing import Callable
 
@@ -29,15 +30,52 @@ def test_from_file_when_not_exists_raises_expected_exception() -> None:
 def test_from_file_when_file_empty_raises_expected_exception(
     tmp_path: Path,
 ) -> None:
+    """
+    Test that creating a DBNStore from an empty file raises a ValueError.
+    """
     # Arrange
     path = tmp_path / "test.dbn"
     path.touch()
 
     # Act, Assert
-    with pytest.raises(RuntimeError):
+    with pytest.raises(ValueError):
         DBNStore.from_file(path)
 
 
+def test_from_file_when_buffer_corrupted_raises_expected_exception(
+    tmp_path: Path,
+) -> None:
+    """
+    Test that creating a DBNStore from an invalid DBN file raises a BentoError.
+    """
+    # Arrange
+    path = tmp_path / "corrupted.dbn"
+    path.write_text("this is a test")
+
+    # Act, Assert
+    with pytest.raises(BentoError):
+        DBNStore.from_file(path)
+
+
+def test_from_bytes_when_buffer_empty_raises_expected_exception() -> None:
+    """
+    Test that creating a DBNStore from an empty buffer raises a ValueError.
+    """
+    # Arrange, Act, Assert
+    with pytest.raises(ValueError):
+        DBNStore.from_bytes(BytesIO())
+
+
+def test_from_bytes_when_buffer_corrupted_raises_expected_exception() -> None:
+    """
+    Test that creating a DBNStore from an invalid DBN stream raises a
+    BentoError.
+    """
+    # Arrange, Act, Assert
+    with pytest.raises(ValueError):
+        DBNStore.from_bytes(BytesIO())
+
+
 def test_sources_metadata_returns_expected_json_as_dict(
     test_data: Callable[[Schema], bytes],
 ) -> None: