Escape formulae on export

src/tablib/formats/_xlsx.py

@@ -35,12 +35,16 @@ def detect(cls, stream):
             return False
 
     @classmethod
-    def export_set(cls, dataset, freeze_panes=True, invalid_char_subst="-"):
+    def export_set(cls, dataset, freeze_panes=True, invalid_char_subst="-", escape=False):
         """Returns XLSX representation of Dataset.
 
         If dataset.title contains characters which are considered invalid for an XLSX file
         sheet name (http://www.excelcodex.com/2012/06/worksheets-naming-conventions/), they will
         be replaced with `invalid_char_subst`.
+
+        If escape is True, formulae will have the leading '=' character removed.
+        This is a security measure to prevent formulae from executing by default
+        in exported XLSX files.
         """
         wb = Workbook()
         ws = wb.worksheets[0]
@@ -50,19 +54,16 @@ def export_set(cls, dataset, freeze_panes=True, invalid_char_subst="-"):
             if dataset.title else 'Tablib Dataset'
         )
 
-        cls.dset_sheet(dataset, ws, freeze_panes=freeze_panes)
+        cls.dset_sheet(dataset, ws, freeze_panes=freeze_panes, escape=escape)
 
         stream = BytesIO()
         wb.save(stream)
         return stream.getvalue()
 
     @classmethod
-    def export_book(cls, databook, freeze_panes=True, invalid_char_subst="-"):
+    def export_book(cls, databook, freeze_panes=True, invalid_char_subst="-", escape=False):
         """Returns XLSX representation of DataBook.
-
-        If dataset.title contains characters which are considered invalid for an XLSX file
-        sheet name (http://www.excelcodex.com/2012/06/worksheets-naming-conventions/), they will
-        be replaced with `invalid_char_subst`.
+        See export_set().
         """
 
         wb = Workbook()
@@ -75,7 +76,7 @@ def export_book(cls, databook, freeze_panes=True, invalid_char_subst="-"):
                 if dset.title else 'Sheet%s' % (i)
             )
 
-            cls.dset_sheet(dset, ws, freeze_panes=freeze_panes)
+            cls.dset_sheet(dset, ws, freeze_panes=freeze_panes, escape=escape)
 
         stream = BytesIO()
         wb.save(stream)
@@ -125,7 +126,7 @@ def import_book(cls, dbook, in_stream, headers=True, read_only=True):
             dbook.add_sheet(data)
 
     @classmethod
-    def dset_sheet(cls, dataset, ws, freeze_panes=True):
+    def dset_sheet(cls, dataset, ws, freeze_panes=True, escape=False):
         """Completes given worksheet from given Dataset."""
         _package = dataset._package(dicts=False)
 
@@ -166,3 +167,6 @@ def dset_sheet(cls, dataset, ws, freeze_panes=True):
                     cell.value = col
                 except (ValueError, TypeError):
                     cell.value = str(col)
+
+                if escape and cell.data_type == 'f' and cell.value.startswith('='):
+                    cell.value = cell.value.replace("=", "")

tests/test_tablib.py

@@ -14,6 +14,7 @@
 from uuid import uuid4
 
 from MarkupPy import markup
+from openpyxl.reader.excel import load_workbook
 
 import tablib
 from tablib.core import Row, detect_format
@@ -1117,6 +1118,68 @@ def test_xlsx_cell_values(self):
             data = tablib.Dataset().load(fh)
         self.assertEqual(data.headers[0], 'Hello World')
 
+    def test_xlsx_export_set_escape_formulae(self):
+        """
+        Test that formulae are sanitised on export.
+        """
+        data.append(('=SUM(1+1)',))
+        _xlsx = data.export('xlsx')
+
+        # read back using openpyxl because tablib reads formulae as values
+        wb = load_workbook(filename=BytesIO(_xlsx))
+        ws = wb.active
+        self.assertEqual('=SUM(1+1)', ws['A1'].value)
+
+        _xlsx = data.export('xlsx', escape=True)
+        wb = load_workbook(filename=BytesIO(_xlsx))
+        ws = wb.active
+        self.assertEqual('SUM(1+1)', ws['A1'].value)
+
+    def test_xlsx_export_book_escape_formulae(self):
+        """
+        Test that formulae are sanitised on export.
+        """
+        data.append(('=SUM(1+1)',))
+        _book = tablib.Databook()
+        _book.add_sheet(data)
+        _xlsx = _book.export('xlsx')
+
+        # read back using openpyxl because tablib reads formulae as values
+        wb = load_workbook(filename=BytesIO(_xlsx))
+        ws = wb.active
+        self.assertEqual('=SUM(1+1)', ws['A1'].value)
+
+        _xlsx = _book.export('xlsx', escape=True)
+        wb = load_workbook(filename=BytesIO(_xlsx))
+        ws = wb.active
+        self.assertEqual('SUM(1+1)', ws['A1'].value)
+
+    def test_xlsx_export_set_escape_formulae_in_header(self):
+        data.headers = ('=SUM(1+1)',)
+        _xlsx = data.export('xlsx')
+        wb = load_workbook(filename=BytesIO(_xlsx))
+        ws = wb.active
+        self.assertEqual('=SUM(1+1)', ws['A1'].value)
+
+        _xlsx = data.export('xlsx', escape=True)
+        wb = load_workbook(filename=BytesIO(_xlsx))
+        ws = wb.active
+        self.assertEqual('SUM(1+1)', ws['A1'].value)
+
+    def test_xlsx_export_book_escape_formulae_in_header(self):
+        data.headers = ('=SUM(1+1)',)
+        _book = tablib.Databook()
+        _book.add_sheet(data)
+        _xlsx = _book.export('xlsx')
+        wb = load_workbook(filename=BytesIO(_xlsx))
+        ws = wb.active
+        self.assertEqual('=SUM(1+1)', ws['A1'].value)
+
+        _xlsx = _book.export('xlsx', escape=True)
+        wb = load_workbook(filename=BytesIO(_xlsx))
+        ws = wb.active
+        self.assertEqual('SUM(1+1)', ws['A1'].value)
+
     def test_xlsx_bad_dimensions(self):
         """Test loading file with bad dimension.  Must be done with
         read_only=False."""