Simplify _read_and_check_tsv_file

clinica/pipelines/statistics_surface/_inputs.py

@@ -26,20 +26,16 @@ def _read_and_check_tsv_file(tsv_file: PathLike) -> pd.DataFrame:
     tsv_data : pd.DataFrame
         DataFrame obtained from the file.
     """
-    if not Path(tsv_file).exists():
-        raise FileNotFoundError(f"File {tsv_file} does not exist.")
-    tsv_data = pd.read_csv(tsv_file, sep="\t")
-    if len(tsv_data.columns) < 2:
-        raise ValueError(f"The TSV data in {tsv_file} should have at least 2 columns.")
-    if tsv_data.columns[0] != TSV_FIRST_COLUMN:
-        raise ValueError(
-            f"The first column in {tsv_file} should always be {TSV_FIRST_COLUMN}."
+    try:
+        return pd.read_csv(tsv_file, sep="\t").set_index(
+            [TSV_FIRST_COLUMN, TSV_SECOND_COLUMN]
         )
-    if tsv_data.columns[1] != TSV_SECOND_COLUMN:
+    except FileNotFoundError:
+        raise FileNotFoundError(f"File {tsv_file} does not exist.")
+    except KeyError:
         raise ValueError(
-            f"The second column in {tsv_file} should always be {TSV_SECOND_COLUMN}."
+            f"The TSV data should have at least two columns: {TSV_FIRST_COLUMN} and {TSV_SECOND_COLUMN}"
         )
-    return tsv_data
 
 
 def _get_t1_freesurfer_custom_file_template(base_dir: PathLike) -> str:

test/unittests/pipelines/statistics_surface/test_inputs.py

@@ -7,30 +7,31 @@
 CURRENT_DIR = os.path.dirname(os.path.realpath(__file__))
 
 
-def test_read_and_check_tsv_file(tmpdir):
+def test_read_and_check_tsv_file_filenotfound_error(tmpdir):
     from clinica.pipelines.statistics_surface._inputs import _read_and_check_tsv_file
 
     with pytest.raises(FileNotFoundError, match="File foo.tsv does not exist"):
         _read_and_check_tsv_file(Path("foo.tsv"))
-    df = pd.DataFrame(columns=["foo"])
-    df.to_csv(tmpdir / "foo.tsv", sep="\t", index=False)
-    with pytest.raises(
-        ValueError, match=r"The TSV data in .*foo.tsv should have at least 2 columns."
-    ):
-        _read_and_check_tsv_file(tmpdir / "foo.tsv")
-    df = pd.DataFrame(columns=["foo", "bar"])
+
+
+@pytest.mark.parametrize(
+    "columns", [["foo"], ["foo", "bar"], ["participant_id", "bar"]]
+)
+def test_read_and_check_tsv_file_data_errors(tmpdir, columns):
+    from clinica.pipelines.statistics_surface._inputs import _read_and_check_tsv_file
+
+    df = pd.DataFrame(columns=columns)
     df.to_csv(tmpdir / "foo.tsv", sep="\t", index=False)
     with pytest.raises(
         ValueError,
-        match=r"The first column in .*foo.tsv should always be participant_id.",
-    ):
-        _read_and_check_tsv_file(tmpdir / "foo.tsv")
-    df = pd.DataFrame(columns=["participant_id", "bar"])
-    df.to_csv(tmpdir / "foo.tsv", sep="\t", index=False)
-    with pytest.raises(
-        ValueError, match=r"The second column in .*foo.tsv should always be session_id."
+        match=r"The TSV data should have at least two columns: participant_id and session_id",
     ):
         _read_and_check_tsv_file(tmpdir / "foo.tsv")
+
+
+def test_read_and_check_tsv_file():
+    from clinica.pipelines.statistics_surface._inputs import _read_and_check_tsv_file
+
     df = _read_and_check_tsv_file(Path(CURRENT_DIR) / "data/subjects.tsv")
     assert len(df) == 7
-    assert set(df.columns) == {"participant_id", "session_id", "group", "age", "sex"}
+    assert set(df.columns) == {"group", "age", "sex"}