Menu
▾
▴
#38 base64 '=' sign strip in incoming messages
As GoldEd shows or saves a message with base64
encoded file in the body it looses trailing equals signs
(padding) of base64 stream. As a result no decoding
program is able to decode a file properly.
I.e. sent sequence was:
------------59D33D519FDD
Content-Type: application/x-zip-compressed;
name="test.zip"
Content-transfer-encoding: base64
Content-Disposition: attachment; filename="test.zip"
ICA=
------------59D33D519FDD--
while received one is:
------------59D33D519FDD
Content-Type: application/x-zip-compressed;
name="test.zip"
Content-transfer-encoding: base64
Content-Disposition: attachment; filename="test.zip"
ICA
------------59D33D519FDD--
When look into the message source itself (Alt-i) the equal
sign is still there.
Logged In: YES
user_id=1042
You probably should be able to decode this with built-in
decoder (Alt-X by default).
Logged In: YES
user_id=433091
As I tried to decode that with embedded decoder I get the
broken file.
Actually, the problem is in the STRIPHTML ON clause. After ^H
I could properly view, save and decode such a file. So that
must be an error in the html stripper code.
Logged In: YES
user_id=1042
This is a feature! STRIPHTML activates QP decoding which in
it's turn cause elimination of (some) equal signs.
Logged In: YES
user_id=433091
That is a strange feature however. Descoder (Alt-X) should
work in any case, I suppose. And if I understand it right equal
signs in QP are only eliminated if followed up by the characters
that may be converted using QP. Trailing equal signs cannot
be converted in any way so should remain.
Logged In: NO
Trailing equal signs are used to allow wrapping of
QP-encoded mail. Equal sign followed by linefeed should be
ignored together with the following linefeed. F.e.:
=E3=E4=E5=E6=
=E7=E8=E9=EA
is the single word and the same as:
=E3=E4=E5=E6=E7=E8=E9=EA
but not the same as:
=E3=E4=E5=E6
=E7=E8=E9=EA
or:
=E3=E4=E5=E6 =E7=E8=E9=EA
which are two separate words.
Logged In: YES
user_id=433091
Well, but what do you think about internal decoder which fails
to decode the file?