Issue #22088: Clarify base-64 alphabets and which characters are discarded

* There are only two base-64 alphabets defined by the RFCs, not three
* Due to the internal translation, plus (+) and slash (/) are never discarded
* standard_ and urlsafe_b64decode() discard characters as well

2 files changed, 27 insertions, 13 deletions

diff --git a/Lib/base64.py b/Lib/base64.py
index 844907f..38bc61e 100755
--- a/Lib/base64.py
+++ b/Lib/base64.py
@@ -65,8 +65,9 @@ def b64decode(s, altchars=None):
     alternative alphabet used instead of the '+' and '/' characters.
 
     The decoded string is returned.  A TypeError is raised if s is
-    incorrectly padded.  Non-base64-alphabet characters are discarded prior
-    to the padding check.
+    incorrectly padded.  Characters that are neither in the normal base-64
+    alphabet nor the alternative alphabet are discarded prior to the padding
+    check.
     """
     if altchars is not None:
         s = s.translate(string.maketrans(altchars[:2], '+/'))
@@ -87,9 +88,10 @@ def standard_b64encode(s):
 def standard_b64decode(s):
     """Decode a string encoded with the standard Base64 alphabet.
 
-    s is the string to decode.  The decoded string is returned.  A TypeError
-    is raised if the string is incorrectly padded or if there are non-alphabet
-    characters present in the string.
+    Argument s is the string to decode.  The decoded string is returned.  A
+    TypeError is raised if the string is incorrectly padded.  Characters that
+    are not in the standard alphabet are discarded prior to the padding
+    check.
     """
     return b64decode(s)
 
@@ -97,19 +99,20 @@ _urlsafe_encode_translation = string.maketrans(b'+/', b'-_')
 _urlsafe_decode_translation = string.maketrans(b'-_', b'+/')
 
 def urlsafe_b64encode(s):
-    """Encode a string using a url-safe Base64 alphabet.
+    """Encode a string using the URL- and filesystem-safe Base64 alphabet.
 
-    s is the string to encode.  The encoded string is returned.  The alphabet
-    uses '-' instead of '+' and '_' instead of '/'.
+    Argument s is the string to encode.  The encoded string is returned.  The
+    alphabet uses '-' instead of '+' and '_' instead of '/'.
     """
     return b64encode(s).translate(_urlsafe_encode_translation)
 
 def urlsafe_b64decode(s):
-    """Decode a string encoded with the standard Base64 alphabet.
+    """Decode a string using the URL- and filesystem-safe Base64 alphabet.
 
-    s is the string to decode.  The decoded string is returned.  A TypeError
-    is raised if the string is incorrectly padded or if there are non-alphabet
-    characters present in the string.
+    Argument s is the string to decode.  The decoded string is returned.  A
+    TypeError is raised if the string is incorrectly padded.  Characters that
+    are not in the URL-safe base-64 alphabet, and are not a plus '+' or slash
+    '/', are discarded prior to the padding check.
 
     The alphabet uses '-' instead of '+' and '_' instead of '/'.
     """
@@ -267,7 +270,7 @@ def b16decode(s, casefold=False):
     a lowercase alphabet is acceptable as input.  For security purposes, the
     default is False.
 
-    The decoded string is returned.  A TypeError is raised if s were
+    The decoded string is returned.  A TypeError is raised if s is
     incorrectly padded or if there are non-alphabet characters present in the
     string.
     """
diff --git a/Lib/test/test_base64.py b/Lib/test/test_base64.py
index 5dd283b..6e67dc0 100644
--- a/Lib/test/test_base64.py
+++ b/Lib/test/test_base64.py
@@ -153,6 +153,13 @@ class BaseXYTestCase(unittest.TestCase):
                  (b'YWJj\nYWI=', b'abcab'))
         for bstr, res in tests:
             self.assertEqual(base64.b64decode(bstr), res)
+            self.assertEqual(base64.standard_b64decode(bstr), res)
+            self.assertEqual(base64.urlsafe_b64decode(bstr), res)
+
+        # Normal alphabet characters not discarded when alternative given
+        res = b'\xFB\xEF\xBE\xFF\xFF\xFF'
+        self.assertEqual(base64.b64decode(b'++[[//]]', b'[]'), res)
+        self.assertEqual(base64.urlsafe_b64decode(b'++--//__'), res)
 
     def test_b32encode(self):
         eq = self.assertEqual
@@ -220,6 +227,10 @@ class BaseXYTestCase(unittest.TestCase):
         eq(base64.b16decode('0102abcdef', True), '\x01\x02\xab\xcd\xef')
         # Non-bytes
         eq(base64.b16decode(bytearray("0102ABCDEF")), '\x01\x02\xab\xcd\xef')
+        # Non-alphabet characters
+        self.assertRaises(TypeError, base64.b16decode, '0102AG')
+        # Incorrect "padding"
+        self.assertRaises(TypeError, base64.b16decode, '010')