Fix SpectraMax kinetic time parsing when edge wells are skipped

lambda/src/data_hub_lambda/spectramax_plate_reader/utils.py

@@ -362,11 +362,13 @@ def parse_raw_well_data(file_path: Path) -> pd.DataFrame:
                 for row_idx in range(num_rows):
                     row_fields = lines[i].split("\t")
 
-                    if row_idx == 0:
-                        time_str = row_fields[0].strip()
-                        time_val = time_str if time_str else None
-                        temp_str = row_fields[1].strip()
-                        temp_val = float(temp_str) if temp_str else None
+                    # SoftMax Pro writes elapsed time / temperature on the first
+                    # populated row of each reading group. When leading rows are
+                    # unselected (e.g. edge wells skipped), that is not row A.
+                    if time_val is None and row_fields and row_fields[0].strip():
+                        time_val = row_fields[0].strip()
+                    if temp_val is None and len(row_fields) > 1 and row_fields[1].strip():
+                        temp_val = float(row_fields[1].strip())
 
                     row_label = _ROW_LABELS[row_idx]
 

lambda/tests/spectramax_plate_reader/test_parse_metadata.py

@@ -112,6 +112,16 @@ def test_kinetic_absorbance(self) -> None:
             "wavelengths": ["595"],
         }
 
+    def test_kinetic_edge_wells_skipped(self) -> None:
+        result = parse_metadata(
+            _FIXTURES_DIR / "spectramax_plate_reader_kinetic_edge_wells_skipped.xls"
+        )
+        assert result == {
+            "measurement_mode": "Absorbance",
+            "measurement_type": "Kinetic",
+            "wavelengths": ["595"],
+        }
+
     def test_endpoint_flat(self) -> None:
         result = parse_metadata(_FIXTURES_DIR / "spectramax_plate_reader_endpoint_flat.xls")
         assert result == {

lambda/tests/spectramax_plate_reader/test_parse_raw_well_data.py

@@ -239,6 +239,57 @@ def test_plate5_row_count(self) -> None:
         assert len(self.df[self.df["plate_name"] == "Plate5"]) == 241 * 96
 
 
+class TestKineticEdgeWellsSkipped:
+    """Kinetic / Absorbance / 595 nm with edge wells unselected.
+
+    SoftMax Pro leaves rows A/H and columns 1/12 empty and writes elapsed
+    time + temperature on the first populated row (B), not row A. The parser
+    must still attach those fields to every well in the reading group.
+    """
+
+    @pytest.fixture(autouse=True)
+    def _load(self) -> None:
+        self.df = parse_raw_well_data(
+            _FIXTURES_DIR / "spectramax_plate_reader_kinetic_edge_wells_skipped.xls"
+        )
+
+    def test_columns(self) -> None:
+        assert list(self.df.columns) == _EXPECTED_COLUMNS
+
+    def test_shape(self) -> None:
+        # 2 time points × 6 rows (B–G) × 9 cols (2–10)
+        assert self.df.shape == (108, 8)
+
+    def test_time_points(self) -> None:
+        times = self.df["time"].unique().tolist()
+        assert times == ["00:00:00", "00:15:01"]
+
+    def test_time_and_temperature_not_null(self) -> None:
+        assert bool(self.df["time"].notna().all())
+        assert bool(self.df["temperature_c"].notna().all())
+
+    def test_wells_exclude_edges(self) -> None:
+        wells = set(self.df["well_position"])
+        assert "A1" not in wells
+        assert "H12" not in wells
+        assert wells == {f"{r}{c}" for r in "BCDEFG" for c in range(2, 11)}
+
+    def test_first_well(self) -> None:
+        first = self.df.iloc[0]
+        assert first["time"] == "00:00:00"
+        assert first["well_position"] == "B2"
+        assert first["temperature_c"] == pytest.approx(30.2)
+        assert first["value"] == pytest.approx(0.1020)
+
+    def test_second_time_point(self) -> None:
+        t1 = self.df[self.df["time"] == "00:15:01"]
+        assert len(t1) == 54
+        first = t1.iloc[0]
+        assert first["well_position"] == "B2"
+        assert first["temperature_c"] == pytest.approx(30.0)
+        assert first["value"] == pytest.approx(0.1120)
+
+
 class TestEndpointFlat:
     """Endpoint / Absorbance / 595 nm — flat layout where all 96 wells appear
     on a single data line with well-position column headers (A1, A2, …, H12).