#20 Chars inserted before accents (fix incl)
Here I provide a fix for the problem that mailcrypt
inserts \201 chars for
latin-1 characters when decrypting the buffer.
The problem was that the process used to read stdout
from GPG uses a coding
system that does that, insert marker characters
before accented characters. I
don't know why it does that, but using raw-text seems
to fix it.
Using raw-text makes sense also because if you modify
text in the emacs buffer
before signing, the signature will probably not be
valid.
I'm not sure if using raw-text encoding is correct to
use in all possible cases,
but my one-line patch below seems to fix the problem
I have. I have not checked
if it fixes the signature problems mentioned in the
mailing-list, but I suspect
it could.
Please let me know if you decide to include the fix.
--
M. <blais@iro.umontreal.ca>
dreper:~/conf/.../elisp/mailcrypt-3.5.6$ cvs diff
--unified mc-gpg.el
Index: mc-gpg.el
===================================================================
RCS file:
/u/blais/cvsroot/conf/share/elisp/mailcrypt-3.5.6/mc-gpg.el,v
retrieving revision 1.1
diff --unified -r1.1 mc-gpg.el
--- mc-gpg.el 2001/12/11 21:58:11 1.1
+++ mc-gpg.el 2001/12/22 19:54:21
@@ -190,6 +190,8 @@
(setq proc
(apply 'start-process-shell-command
"*GPG*" mybuf
program args))
+ (set-process-coding-system proc 'raw-text
'raw-text)
+
;; send in passwd if necessary
(if passwd
(progn
Logged In: NO
The same bug exits with pgp2.6.3 using Gnus 5.9.0 as MUA.
.