pythongh-144001: Simplify Base64 decoding with altchars and ignorechars specified (pythonGH-144324)

Treat "+" and "/" like other characters not in the alternative Base64
alphabet when both altchars and ignorechars are specified.
E.g. discard them if they are not in altchars but are in ignorechars,
and set error if they are not in altchars and not in ignorechars.
Only emit warnings if ignorechars is not specified.

Author: serhiy-storchaka

Doc/library/base64.rst

@@ -91,7 +91,8 @@ POST request.
    ``False`` otherwise.
 
    If *validate* is false, characters that are neither
-   in the normal base-64 alphabet nor the alternative alphabet are
+   in the normal base-64 alphabet nor (if *ignorechars* is not specified)
+   the alternative alphabet are
    discarded prior to the padding check, but the ``+`` and ``/`` characters
    keep their meaning if they are not in *altchars* (they will be discarded
    in future Python versions).
@@ -101,15 +102,14 @@ POST request.
 
    For more information about the strict base64 check, see :func:`binascii.a2b_base64`
 
+   .. versionchanged:: next
+      Added the *ignorechars* parameter.
+
    .. deprecated:: next
       Accepting the ``+`` and ``/`` characters with an alternative alphabet
       is now deprecated.
 
 
-   .. versionchanged:: next
-      Added the *ignorechars* parameter.
-
-
 .. function:: standard_b64encode(s)
 
    Encode :term:`bytes-like object` *s* using the standard Base64 alphabet

Lib/base64.py

@@ -89,49 +89,37 @@ def b64decode(s, altchars=None, validate=_NOT_SPECIFIED, *, ignorechars=_NOT_SPE
     s = _bytes_from_decode_data(s)
     if validate is _NOT_SPECIFIED:
         validate = ignorechars is not _NOT_SPECIFIED
-    if ignorechars is _NOT_SPECIFIED:
-        ignorechars = b''
     badchar = None
-    badchar_strict = False
     if altchars is not None:
         altchars = _bytes_from_decode_data(altchars)
         if len(altchars) != 2:
             raise ValueError(f'invalid altchars: {altchars!r}')
-        for b in b'+/':
-            if b not in altchars and b in s:
-                if badchar is None:
-                    badchar = b
-                if not validate:
-                    break
-                if not isinstance(ignorechars, (bytes, bytearray)):
-                    ignorechars = memoryview(ignorechars).cast('B')
-                if b not in ignorechars:
-                    badchar_strict = True
+        if ignorechars is _NOT_SPECIFIED:
+            for b in b'+/':
+                if b not in altchars and b in s:
                     badchar = b
                     break
-        s = s.translate(bytes.maketrans(altchars, b'+/'))
+            s = s.translate(bytes.maketrans(altchars, b'+/'))
+        else:
+            trans = bytes.maketrans(b'+/' + altchars, altchars + b'+/')
+            s = s.translate(trans)
+            ignorechars = ignorechars.translate(trans)
+    if ignorechars is _NOT_SPECIFIED:
+        ignorechars = b''
     result = binascii.a2b_base64(s, strict_mode=validate,
                                  ignorechars=ignorechars)
     if badchar is not None:
         import warnings
-        if badchar_strict:
+        if validate:
             warnings.warn(f'invalid character {chr(badchar)!a} in Base64 data '
                           f'with altchars={altchars!r} and validate=True '
                           f'will be an error in future Python versions',
                           DeprecationWarning, stacklevel=2)
         else:
-            ignorechars = bytes(ignorechars)
-            if ignorechars:
-                warnings.warn(f'invalid character {chr(badchar)!a} in Base64 data '
-                              f'with altchars={altchars!r} '
-                              f'and ignorechars={ignorechars!r} '
-                              f'will be discarded in future Python versions',
-                              FutureWarning, stacklevel=2)
-            else:
-                warnings.warn(f'invalid character {chr(badchar)!a} in Base64 data '
-                              f'with altchars={altchars!r} and validate=False '
-                              f'will be discarded in future Python versions',
-                              FutureWarning, stacklevel=2)
+            warnings.warn(f'invalid character {chr(badchar)!a} in Base64 data '
+                          f'with altchars={altchars!r} and validate=False '
+                          f'will be discarded in future Python versions',
+                          FutureWarning, stacklevel=2)
     return result
 
 

Lib/test/test_base64.py

@@ -291,6 +291,7 @@ def test_b64decode_altchars(self):
             eq(base64.b64decode(data_str, altchars=altchars), res)
             eq(base64.b64decode(data, altchars=altchars_str), res)
             eq(base64.b64decode(data_str, altchars=altchars_str), res)
+            eq(base64.b64decode(data, altchars=altchars, ignorechars=b'\n'), res)
 
         self.assertRaises(ValueError, base64.b64decode, b'', altchars=b'+')
         self.assertRaises(ValueError, base64.b64decode, b'', altchars=b'+/-')
@@ -370,26 +371,28 @@ def test_b64decode_invalid_chars(self):
         self.assertEqual(r, b'\xff\xff\xff')
         self.assertEqual(str(cm.warning),
                          error % ('/', "altchars=b'-_' and validate=True"))
-        with self.assertWarns(FutureWarning) as cm:
-            r = base64.b64decode(b'++++', altchars=b'-_', ignorechars=b'+')
-        self.assertEqual(r, b'\xfb\xef\xbe')
-        self.assertEqual(str(cm.warning),
-                         discarded % ('+', "altchars=b'-_' and ignorechars=b'+'"))
-        with self.assertWarns(FutureWarning) as cm:
-            r = base64.b64decode(b'////', altchars=b'-_', ignorechars=b'/')
-        self.assertEqual(r, b'\xff\xff\xff')
-        self.assertEqual(str(cm.warning),
-                         discarded % ('/', "altchars=b'-_' and ignorechars=b'/'"))
-        with self.assertWarns(DeprecationWarning) as cm:
-            r = base64.b64decode(b'++++////', altchars=b'-_', ignorechars=b'+')
+        r = base64.b64decode(b'++++', altchars=b'-_', ignorechars=b'+')
+        self.assertEqual(r, b'')
+        r = base64.b64decode(b'////', altchars=b'-_', ignorechars=b'/')
+        self.assertEqual(r, b'')
+        r = base64.b64decode(b'++++////', altchars=b'-_', validate=False, ignorechars=b'')
+        self.assertEqual(r, b'')
+        with self.assertRaisesRegex(binascii.Error, 'Only base64 data is allowed'):
+            base64.b64decode(b'////', altchars=b'-_', ignorechars=b'')
+        with self.assertRaisesRegex(binascii.Error, 'Only base64 data is allowed'):
+            base64.b64decode(b'++++', altchars=b'-_', ignorechars=b'')
+        r = base64.b64decode(b'++++YWJj----____', altchars=b'-_', ignorechars=b'+')
+        self.assertEqual(r, b'abc\xfb\xef\xbe\xff\xff\xff')
+        r = base64.b64decode(b'////YWJj----____', altchars=b'-_', ignorechars=b'/')
+        self.assertEqual(r, b'abc\xfb\xef\xbe\xff\xff\xff')
+        r = base64.b64decode(b'++++,,,,', altchars=b'+,', ignorechars=b'+')
         self.assertEqual(r, b'\xfb\xef\xbe\xff\xff\xff')
-        self.assertEqual(str(cm.warning),
-                         error % ('/', "altchars=b'-_' and validate=True"))
-        with self.assertWarns(DeprecationWarning) as cm:
-            r = base64.b64decode(b'++++////', altchars=b'-_', ignorechars=b'/')
+        r = base64.b64decode(b'////YWJj++++,,,,', altchars=b'+,', ignorechars=b'/')
+        self.assertEqual(r, b'abc\xfb\xef\xbe\xff\xff\xff')
+        r = base64.b64decode(b'----////', altchars=b'-/', ignorechars=b'/')
         self.assertEqual(r, b'\xfb\xef\xbe\xff\xff\xff')
-        self.assertEqual(str(cm.warning),
-                         error % ('+', "altchars=b'-_' and validate=True"))
+        r = base64.b64decode(b'++++YWJj----////', altchars=b'-/', ignorechars=b'+')
+        self.assertEqual(r, b'abc\xfb\xef\xbe\xff\xff\xff')
 
         with self.assertWarns(FutureWarning) as cm:
             self.assertEqual(base64.urlsafe_b64decode(b'++++'), b'\xfb\xef\xbe')