fix: xlsx_errors column overwriting last data field in XLSX export (#2194)

Signed-off-by: tdruez <tdruez@aboutcode.org>

Author: tdruez

scanpipe/pipes/output.py

@@ -382,7 +382,7 @@ def add_xlsx_worksheet(workbook, worksheet_name, rows, fields):
     worksheet.write_row(row=0, col=0, data=header, cell_format=cell_format)
 
     errors_count = 0
-    errors_col_index = len(fields) - 1  # rows and cols are zero-indexed
+    errors_col_index = len(fields)  # rows and cols are zero-indexed
 
     for row_index, record in enumerate(rows, start=1):
         row_errors = []

scanpipe/tests/pipes/test_output.py

@@ -839,6 +839,28 @@ def test_add_xlsx_worksheet_does_not_munge_long_strings_of_over_1024_lines(self)
             if r != x:
                 self.assertEqual(r[-50:], x)
 
+    def test_add_xlsx_worksheet_writes_errors_to_dedicated_column(self):
+        rows = [{"foo": "some value", "bar": "f" * 40000}]
+
+        test_dir = Path(tempfile.mkdtemp(prefix="scancode-io-test"))
+        output_file = test_dir / "errors-column.xlsx"
+        with xlsxwriter.Workbook(str(output_file)) as workbook:
+            output.add_xlsx_worksheet(
+                workbook=workbook,
+                worksheet_name="packages",
+                rows=rows,
+                fields=["foo", "bar"],
+            )
+
+        workbook = openpyxl.load_workbook(output_file, read_only=True, data_only=True)
+        worksheet = workbook["packages"]
+        header_row, data_row = list(worksheet.iter_rows(values_only=True))
+
+        self.assertEqual(("foo", "bar", "xlsx_errors"), header_row)
+        self.assertEqual("some value", data_row[0])
+        self.assertEqual("f" * 32767, data_row[1])
+        self.assertIn("has been truncated", data_row[2])
+
     def test__adapt_value_for_xlsx_does_adapt(self):
         result, error = output._adapt_value_for_xlsx(
             fieldname="foo",