gh-118314: Fix padding edge case in binascii.a2b_base64 strict mode (GH-118320)

Fix an edge case in `binascii.a2b_base64` strict mode, where
excessive padding was not detected when no padding is necessary.

Co-authored-by: Terry Jan Reedy <tjreedy@udel.edu>
Co-authored-by: Pieter Eendebak <pieter.eendebak@gmail.com>

diff --git a/Lib/test/test_binascii.py b/Lib/test/test_binascii.py
index 82dea8a6d731ea3b691bc445b755f40dd73e2ee2..1f3b6746ce4a626687d32a3a5c219036925049ef 100644 (file)
--- a/Lib/test/test_binascii.py
+++ b/Lib/test/test_binascii.py
@@ -139,13 +139,21 @@ class BinASCIITest(unittest.TestCase):
         def assertDiscontinuousPadding(data, non_strict_mode_expected_result: bytes):
             _assertRegexTemplate(r'(?i)Discontinuous padding', data, non_strict_mode_expected_result)
 
+        def assertExcessPadding(data, non_strict_mode_expected_result: bytes):
+            _assertRegexTemplate(r'(?i)Excess padding', data, non_strict_mode_expected_result)
+
         # Test excess data exceptions
         assertExcessData(b'ab==a', b'i')
         assertExcessData(b'ab===', b'i')
+        assertExcessData(b'ab====', b'i')
         assertExcessData(b'ab==:', b'i')
         assertExcessData(b'abc=a', b'i\xb7')
         assertExcessData(b'abc=:', b'i\xb7')
         assertExcessData(b'ab==\n', b'i')
+        assertExcessData(b'abc==', b'i\xb7')
+        assertExcessData(b'abc===', b'i\xb7')
+        assertExcessData(b'abc====', b'i\xb7')
+        assertExcessData(b'abc=====', b'i\xb7')
 
         # Test non-base64 data exceptions
         assertNonBase64Data(b'\nab==', b'i')
@@ -157,8 +165,15 @@ class BinASCIITest(unittest.TestCase):
         assertLeadingPadding(b'=', b'')
         assertLeadingPadding(b'==', b'')
         assertLeadingPadding(b'===', b'')
+        assertLeadingPadding(b'====', b'')
+        assertLeadingPadding(b'=====', b'')
         assertDiscontinuousPadding(b'ab=c=', b'i\xb7')
         assertDiscontinuousPadding(b'ab=ab==', b'i\xb6\x9b')
+        assertExcessPadding(b'abcd=', b'i\xb7\x1d')
+        assertExcessPadding(b'abcd==', b'i\xb7\x1d')
+        assertExcessPadding(b'abcd===', b'i\xb7\x1d')
+        assertExcessPadding(b'abcd====', b'i\xb7\x1d')
+        assertExcessPadding(b'abcd=====', b'i\xb7\x1d')
 
 
     def test_base64errors(self):

diff --git a/Misc/NEWS.d/next/Library/2024-04-26-12-42-29.gh-issue-118314.Z7reGc.rst b/Misc/NEWS.d/next/Library/2024-04-26-12-42-29.gh-issue-118314.Z7reGc.rst
new file mode 100644 (file)
index 0000000..ff3ee68
--- /dev/null
+++ b/Misc/NEWS.d/next/Library/2024-04-26-12-42-29.gh-issue-118314.Z7reGc.rst
@@ -0,0 +1 @@
+Fix an edge case in :func:`binascii.a2b_base64` strict mode, where excessive padding is not detected when no padding is necessary.

diff --git a/Modules/binascii.c b/Modules/binascii.c
index 250f03a9531eedc240a74425c3a700353e8c9da4..6bb01d148b6faac9aa981b4368f040927ca176fb 100644 (file)
--- a/Modules/binascii.c
+++ b/Modules/binascii.c
@@ -414,6 +414,13 @@ binascii_a2b_base64_impl(PyObject *module, Py_buffer *data, int strict_mode)
         if (this_ch == BASE64_PAD) {
             padding_started = 1;
 
+            if (strict_mode && quad_pos == 0) {
+                state = get_binascii_state(module);
+                if (state) {
+                    PyErr_SetString(state->Error, "Excess padding not allowed");
+                }
+                goto error_end;
+            }
             if (quad_pos >= 2 && quad_pos + ++pads >= 4) {
                 /* A pad sequence means we should not parse more input.
                 ** We've already interpreted the data from the quad at this point.