Internal cleanups: stale docstring, index_by_time mutation, infer_extension message

src/activity_parser/parser.py

@@ -104,6 +104,8 @@ def infer_extension(file: str | PathLike[str]) -> str:
     ext = p.suffix
     if ext.lower() == ".gz":
         ext = Path(p.stem).suffix
+        if not ext:
+            ext = ".gz"
     return normalize_extension(ext)
 
 

src/activity_parser/postprocess.py

@@ -43,6 +43,7 @@ def index_by_time(records: pd.DataFrame, column: str) -> pd.DataFrame:
     if column not in records.columns:
         if not records.empty:
             return records.rename_axis(column)
+        records = records.copy()
         records[column] = pd.NaT
 
     records = records.set_index(column)

tests/test_parser.py

@@ -34,6 +34,11 @@ def test_infer_extension():
     assert infer_extension("dir/sub/ride.GPX.GZ") == "gpx"
 
 
+def test_infer_extension_bare_gz_rejected():
+    with pytest.raises(ValueError, match="Ambiguous"):
+        infer_extension("ride.gz")
+
+
 def test_select_and_reorder_cols_selects_and_orders():
     df = pd.DataFrame({"b": [1], "a": [2], "ignored": [3]})
     out = select_and_reorder_cols(df, ["a", "b", "missing"], include_all_columns=False)

tests/test_postprocess.py

@@ -43,6 +43,12 @@ def test_index_by_time_missing_column_yields_empty_datetime_index():
     assert isinstance(out.index, pd.DatetimeIndex)
 
 
+def test_index_by_time_missing_column_and_empty_does_not_mutate_input():
+    df = pd.DataFrame({"value": []})
+    index_by_time(df, "time")
+    assert "time" not in df.columns
+
+
 def test_index_by_time_missing_column_with_rows_keeps_them():
     df = pd.DataFrame({"value": [1, 2, 3]})
     out = index_by_time(df, "time")

tests/test_readme.py

@@ -1,4 +1,4 @@
-"""Keeps README.md's tables in sync with output.py."""
+"""Keeps README.md's tables in sync with parser.py."""
 
 import re
 from dataclasses import fields