[pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

Author: pre-commit-ci[bot]

src/core/parquet.py

@@ -12,8 +12,12 @@
 
 def map_arrow_type(arrow_type: pa.DataType) -> FeatureType:
     """Map a PyArrow DataType to an OpenML FeatureType."""
-    if pa.types.is_floating(arrow_type) or pa.types.is_integer(arrow_type) or pa.types.is_decimal(
-        arrow_type
+    if (
+        pa.types.is_floating(arrow_type)
+        or pa.types.is_integer(arrow_type)
+        or pa.types.is_decimal(
+            arrow_type,
+        )
     ):
         return FeatureType.NUMERIC
     if pa.types.is_boolean(arrow_type) or pa.types.is_dictionary(arrow_type):
@@ -51,7 +55,7 @@ def read_parquet_metadata(file_bytes: bytes) -> ParquetMeta:
 
     schema = pf.schema_arrow
     num_rows = pf.metadata.num_rows
-    md5 = hashlib.md5(file_bytes, usedforsecurity=False).hexdigest()  # noqa: S324
+    md5 = hashlib.md5(file_bytes, usedforsecurity=False).hexdigest()
 
     # Read full table once to count per-column nulls
     table = pf.read()
@@ -66,7 +70,7 @@ def read_parquet_metadata(file_bytes: bytes) -> ParquetMeta:
                 name=col_name,
                 data_type=map_arrow_type(schema.field(col_name).type),
                 number_of_missing_values=null_count,
-            )
+            ),
         )
 
     return ParquetMeta(

src/core/storage.py

@@ -8,7 +8,7 @@
 import boto3
 from botocore.exceptions import BotoCoreError, ClientError
 
-from config import _load_configuration, _config_file
+from config import _config_file, _load_configuration
 
 if TYPE_CHECKING:
     from pathlib import Path

src/schemas/datasets/upload.py

@@ -8,7 +8,9 @@
 class DatasetUploadMetadata(BaseModel):
     """Metadata provided alongside the uploaded Parquet file."""
 
-    name: str = Field(description="Human-readable name of the dataset.", min_length=1, max_length=256)
+    name: str = Field(
+        description="Human-readable name of the dataset.", min_length=1, max_length=256
+    )
     description: str = Field(description="Description of the dataset.", min_length=1)
     default_target_attribute: str = Field(
         default="",

tests/routers/openml/dataset_upload_test.py

@@ -88,7 +88,9 @@ def test_upload_unauthenticated(api_client_unauthenticated: TestClient) -> None:
 
 
 def test_upload_non_parquet_file(api_client_authenticated: TestClient) -> None:
-    response = _upload(api_client_authenticated, file_bytes=b"col1,col2\n1,2\n", filename="data.csv")
+    response = _upload(
+        api_client_authenticated, file_bytes=b"col1,col2\n1,2\n", filename="data.csv"
+    )
     assert response.status_code == HTTPStatus.BAD_REQUEST
     assert response.json()["detail"]["code"] == "110"
 
@@ -100,7 +102,9 @@ def test_upload_invalid_parquet_bytes(api_client_authenticated: TestClient) -> N
 
 
 def test_upload_invalid_metadata_json(api_client_authenticated: TestClient) -> None:
-    files = {"file": ("iris.parquet", io.BytesIO(_make_parquet_bytes()), "application/octet-stream")}
+    files = {
+        "file": ("iris.parquet", io.BytesIO(_make_parquet_bytes()), "application/octet-stream")
+    }
     data = {"metadata": "NOT VALID JSON {{{"}
     response = api_client_authenticated.post("/datasets/upload", files=files, data=data)
     assert response.status_code == HTTPStatus.UNPROCESSABLE_ENTITY
@@ -129,7 +133,10 @@ def test_upload_minio_failure_returns_500(api_client_authenticated: TestClient)
     file_bytes = _make_parquet_bytes()
 
     with (
-        patch("routers.openml.datasets.upload_to_minio", side_effect=RuntimeError("connection refused")),
+        patch(
+            "routers.openml.datasets.upload_to_minio",
+            side_effect=RuntimeError("connection refused"),
+        ),
         patch("database.datasets.insert_file", return_value=99),
         patch("database.datasets.insert_dataset", return_value=42),
     ):