Bug wrong encoding when uploading Nordic characters through files API (…

…#1398)

CHANGELOG.md

@@ -17,6 +17,11 @@ Changes are grouped as follows
 - `Fixed` for any bug fixes.
 - `Security` in case of vulnerabilities.
 
+## [6.28.3] - 2023-10-03
+### Fixed
+- When uploading files as strings using `client.files.upload_bytes` the wrong encoding is used on Windows, which is causing
+  part of the content to be lost when uploading. This is now fixed.
+
 ## [6.28.2] - 2023-10-02
 ### Fixed
 - When cache lookup did not yield a token for `CredentialProvider`s like `OAuthDeviceCode` or `OAuthInteractive`, a

cognite/client/_api/files.py

@@ -545,6 +545,9 @@ def upload_bytes(
                 >>> c = CogniteClient()
                 >>> res = c.files.upload_bytes(b"some content", name="my_file", asset_ids=[1,2,3])
         """
+        if isinstance(content, str):
+            content = content.encode("utf-8")
+
         file_metadata = FileMetadata(
             name=name,
             external_id=external_id,

cognite/client/_version.py

@@ -1,4 +1,4 @@
 from __future__ import annotations
 
-__version__ = "6.28.2"
+__version__ = "6.28.3"
 __api_subversion__ = "V20220125"

pyproject.toml

@@ -1,7 +1,7 @@
 [tool.poetry]
 name = "cognite-sdk"
 
-version = "6.28.2"
+version = "6.28.3"
 description = "Cognite Python SDK"
 readme = "README.md"
 documentation = "https://cognite-sdk-python.readthedocs-hosted.com"

tests/tests_integration/test_api/test_files.py

@@ -3,6 +3,7 @@
 
 import pytest
 
+from cognite.client import CogniteClient
 from cognite.client.data_classes import (
     FileMetadata,
     FileMetadataFilter,
@@ -202,3 +203,14 @@ def test_filter_file_on_geoLocation(self, cognite_client, new_file_with_geoLocat
             time.sleep(0.2)
             res = cognite_client.files.list(geo_location=geo_location_filter)
         assert res[0].geo_location == new_file_with_geoLocation.geo_location
+
+    def test_upload_bytes_with_nordic_characters(self, cognite_client: CogniteClient) -> None:
+        content = "æøåøøøø ååå ææææ"
+        external_id = "test_upload_bytes_with_nordic_characters"
+
+        _ = cognite_client.files.upload_bytes(
+            content=content, name="nordic_chars.txt", external_id=external_id, overwrite=True
+        )
+
+        retrieved_content = cognite_client.files.download_bytes(external_id=external_id)
+        assert retrieved_content == content.encode("utf-8")