some linting

tests/conftest.py

@@ -44,7 +44,7 @@ def _tfs_complex_file() -> pathlib.Path:
     return INPUTS_DIR / "complex.tfs"
 
 
-@pytest.fixture()
+@pytest.fixture
 def _tfs_madng_file() -> pathlib.Path:
     """A TFS file withboth complex values and booleans, in headers and columns."""
     return INPUTS_DIR / "madng.tfs"

tests/test_reader.py

@@ -1,7 +1,6 @@
 import pathlib
 
 import pytest
-import numpy as np
 from pandas.api import types as pdtypes
 from pandas.core.arrays.string_ import StringDtype
 from pandas.testing import assert_frame_equal

tests/test_validation.py

@@ -27,7 +27,7 @@ def test_warn_unphysical_values_in_headers(self, _tfs_dataframe, validation_mode
         df = _tfs_dataframe
         df.headers["NANVALUE"] = np.nan  # make sure there is a NaN in there
         validate(df, compatibility=validation_mode)
-        
+
         for record in caplog.records:
             assert record.levelname == "WARNING"
         assert "contains non-physical values in headers" in caplog.text

tests/test_writer.py

@@ -1,5 +1,4 @@
 import logging
-import pathlib
 import random
 import string
 
@@ -9,7 +8,6 @@
 from cpymad.madx import Madx
 from pandas._testing import assert_dict_equal
 from pandas.testing import assert_frame_equal, assert_series_equal
-from pymadng import MAD
 
 import tfs
 from tfs import TfsDataFrame, read_tfs, write_tfs
@@ -216,7 +214,7 @@ def test_tfs_write_read_with_complex(self, _tfs_dataframe_complex, tmp_path):
     def test_tfs_write_with_nil_in_headers(self, _tfs_dataframe, tmp_path):
         df = _tfs_dataframe
         df.headers["WRITETONIL"] = None
-        
+
         write_location = tmp_path / "test.tfs"
         write_tfs(write_location, df, validate="madng")
 

tfs/writer.py

@@ -132,7 +132,7 @@ def write_tfs(
     # to floats, and don't try to convert strings to its special StringDtype (as then 'None'
     # would be transformed into <pd.NA> and if we write this to file we are very much cooked.)
     # Overall we do not care to infer specialized dtypes, just that it makes the best inference
-    # to valid dtypes (i.e. an object column should be inferred as strings if that makes sense). 
+    # to valid dtypes (i.e. an object column should be inferred as strings if that makes sense).
     data_frame = data_frame.convert_dtypes(convert_integer=False, convert_floating=False, convert_string=None)
 
     if save_index: