fix: local data source relative path includes the first directory (#702)

Author: speg03

src/sagemaker/local/data.py

@@ -44,7 +44,7 @@ def get_data_source_instance(data_source, sagemaker_session):
     """
     parsed_uri = urlparse(data_source)
     if parsed_uri.scheme == 'file':
-        return LocalFileDataSource(parsed_uri.path)
+        return LocalFileDataSource(parsed_uri.netloc + parsed_uri.path)
     elif parsed_uri.scheme == 's3':
         return S3DataSource(parsed_uri.netloc, parsed_uri.path, sagemaker_session)
 

tests/unit/test_local_data.py

@@ -28,6 +28,10 @@ def test_get_data_source_instance_with_file(LocalFileDataSource, sagemaker_local
     LocalFileDataSource.assert_called_with('/my/file')
     assert data_source is not None
 
+    data_source = sagemaker.local.data.get_data_source_instance('file://relative/path', sagemaker_local_session)
+    LocalFileDataSource.assert_called_with('relative/path')
+    assert data_source is not None
+
 
 @patch('sagemaker.local.data.S3DataSource')
 def test_get_data_source_instance_with_s3(S3DataSource, sagemaker_local_session):