gh-115801: Only allow sequence of strings as input for difflib.unified_diff (GH-118333)

diff --git a/Lib/difflib.py b/Lib/difflib.py
index 0443963b4fd6979ff5da600aef3d415832c72a45..7f595b6c72e641df6aa5cf18b0d2093f72992a56 100644 (file)
--- a/Lib/difflib.py
+++ b/Lib/difflib.py
@@ -1264,6 +1264,12 @@ def _check_types(a, b, *args):
     if b and not isinstance(b[0], str):
         raise TypeError('lines to compare must be str, not %s (%r)' %
                         (type(b[0]).__name__, b[0]))
+    if isinstance(a, str):
+        raise TypeError('input must be a sequence of strings, not %s' %
+                        type(a).__name__)
+    if isinstance(b, str):
+        raise TypeError('input must be a sequence of strings, not %s' %
+                        type(b).__name__)
     for arg in args:
         if not isinstance(arg, str):
             raise TypeError('all arguments must be str, not: %r' % (arg,))

diff --git a/Lib/test/test_difflib.py b/Lib/test/test_difflib.py
index bf6e5b1152b4a2b18bbeae425b474bdfeb587c4f..9e217249be7332d00197ac905a9c513f43d2cd90 100644 (file)
--- a/Lib/test/test_difflib.py
+++ b/Lib/test/test_difflib.py
@@ -295,7 +295,7 @@ class TestDiffer(unittest.TestCase):
 
 class TestOutputFormat(unittest.TestCase):
     def test_tab_delimiter(self):
-        args = ['one', 'two', 'Original', 'Current',
+        args = [['one'], ['two'], 'Original', 'Current',
             '2005-01-26 23:30:50', '2010-04-02 10:20:52']
         ud = difflib.unified_diff(*args, lineterm='')
         self.assertEqual(list(ud)[0:2], [
@@ -307,7 +307,7 @@ class TestOutputFormat(unittest.TestCase):
                            "--- Current\t2010-04-02 10:20:52"])
 
     def test_no_trailing_tab_on_empty_filedate(self):
-        args = ['one', 'two', 'Original', 'Current']
+        args = [['one'], ['two'], 'Original', 'Current']
         ud = difflib.unified_diff(*args, lineterm='')
         self.assertEqual(list(ud)[0:2], ["--- Original", "+++ Current"])
 
@@ -447,6 +447,28 @@ class TestBytes(unittest.TestCase):
                                     lineterm=b'')
         assertDiff(expect, actual)
 
+
+class TestInputTypes(unittest.TestCase):
+    def _assert_type_error(self, msg, generator, *args):
+        with self.assertRaises(TypeError) as ctx:
+            list(generator(*args))
+        self.assertEqual(msg, str(ctx.exception))
+
+    def test_input_type_checks(self):
+        unified = difflib.unified_diff
+        context = difflib.context_diff
+
+        expect = "input must be a sequence of strings, not str"
+        self._assert_type_error(expect, unified, 'a', ['b'])
+        self._assert_type_error(expect, context, 'a', ['b'])
+
+        self._assert_type_error(expect, unified, ['a'], 'b')
+        self._assert_type_error(expect, context, ['a'], 'b')
+
+        expect = "lines to compare must be str, not NoneType (None)"
+        self._assert_type_error(expect, unified, ['a'], [None])
+        self._assert_type_error(expect, context, ['a'], [None])
+
     def test_mixed_types_content(self):
         # type of input content must be consistent: all str or all bytes
         a = [b'hello']
@@ -495,10 +517,6 @@ class TestBytes(unittest.TestCase):
         b = ['bar\n']
         list(difflib.unified_diff(a, b, 'a', 'b', datea, dateb))
 
-    def _assert_type_error(self, msg, generator, *args):
-        with self.assertRaises(TypeError) as ctx:
-            list(generator(*args))
-        self.assertEqual(msg, str(ctx.exception))
 
 class TestJunkAPIs(unittest.TestCase):
     def test_is_line_junk_true(self):

diff --git a/Misc/NEWS.d/next/Core and Builtins/2024-04-27-18-36-46.gh-issue-115801.SVeHSy.rst b/Misc/NEWS.d/next/Core and Builtins/2024-04-27-18-36-46.gh-issue-115801.SVeHSy.rst
new file mode 100644 (file)
index 0000000..93b176d
--- /dev/null
+++ b/Misc/NEWS.d/next/Core and Builtins/2024-04-27-18-36-46.gh-issue-115801.SVeHSy.rst
@@ -0,0 +1 @@
+Raise ``TypeError`` when passing a string to :func:`difflib.unified_diff` and :func:`difflib.context_diff`.