Worksheet.iter_rows() with empty sheet: TypeError: 'Worksheet' object is not an iterator

The method Worksheet.iter_rows() is documented as returning a generator.

In my code, I call next() on this returned value which should work if the returned value is a generator (or iterator).

However, when the worksheet is empty, a tuple is returned. See attached patch and test for fix.

diff -r 153497a0dc2b openpyxl/worksheet/tests/test_worksheet.py
--- a/openpyxl/worksheet/tests/test_worksheet.py        Wed Jun 24 19:10:27 2020 +0200
+++ b/openpyxl/worksheet/tests/test_worksheet.py        Tue Jun 30 13:13:13 2020 -0700
@@ -3,6 +3,7 @@
 # test imports
 import pytest
 
+import collections.abc
 from itertools import islice
 
 # package imports
@@ -87,6 +88,12 @@
         for row, coord in zip(rows, expected):
             assert tuple(c.coordinate for c in row) == coord
 
+    def test_iter_rows_is_iterator(self, Worksheet):
+        ws = Worksheet(Workbook())
+        it = ws.iter_rows()
+        assert isinstance(it, collections.abc.Iterator)
+        with pytest.raises(StopIteration):
+            next(it)
 
     def test_cell_alternate_coordinates(self, Worksheet):
         ws = Worksheet(Workbook())
@@ -208,7 +215,7 @@
 
     def test_no_rows(self, Worksheet):
         ws = Worksheet(Workbook())
-        assert ws.rows == ()
+        assert tuple(ws.rows) == ()
 
 
     def test_no_cols(self, Worksheet):
diff -r 153497a0dc2b openpyxl/worksheet/worksheet.py
--- a/openpyxl/worksheet/worksheet.py   Wed Jun 24 19:10:27 2020 +0200
+++ b/openpyxl/worksheet/worksheet.py   Tue Jun 30 13:13:13 2020 -0700
@@ -429,15 +429,12 @@
         :rtype: generator
         """
 
-        if self._current_row == 0 and not any([min_col, min_row, max_col, max_row ]):
-            return ()
-
-        min_col = min_col or 1
-        min_row = min_row or 1
-        max_col = max_col or self.max_column
-        max_row = max_row or self.max_row
-
-        return self._cells_by_row(min_col, min_row, max_col, max_row, values_only)
+        if self._current_row != 0 or any([min_col, min_row, max_col, max_row]):
+            min_col = min_col or 1
+            min_row = min_row or 1
+            max_col = max_col or self.max_column
+            max_row = max_row or self.max_row
+            yield from self._cells_by_row(min_col, min_row, max_col, max_row, values_only)