GPG --status-fd outputs in UTF-8.

GPG --status-fd outputs in UTF-8.

[Michaël Cadilhac]

[-- Attachment #1.1: Type: text/plain, Size: 539 bytes --]

Hi!

It can be read from the DETAILS file of Gnupg the following :

,----
| Format of the "--status-fd" output
| ==================================
| Every line is prefixed with "[GNUPG:] ", followed by a keyword with
| the type of the status line and a some arguments depending on the
| type (maybe none)[...]
|         The username is the primary one encoded in UTF-8
`----

To be consistent with the rest of gpg(1)'s output, it would be good to
ask Gnupg to output everything in UTF-8 and treat the buffer as such.

Here's a proposal :

[-- Attachment #1.2: pgg-gpg.patch --]
[-- Type: text/x-patch, Size: 1466 bytes --]

Index: lisp/pgg-gpg.el
===================================================================
RCS file: /usr/local/cvsroot/gnus/lisp/pgg-gpg.el,v
retrieving revision 7.27
diff -B -w -c -r7.27 pgg-gpg.el
*** lisp/pgg-gpg.el	1 Mar 2007 23:43:33 -0000	7.27
--- lisp/pgg-gpg.el	7 Mar 2007 18:09:13 -0000
***************
*** 72,77 ****
--- 72,78 ----
    (let* ((output-file-name (pgg-make-temp-file "pgg-output"))
  	 (args
  	  (append (list "--no-tty"
+ 			"--display-charset" "utf-8"
  			"--status-fd" "1"
  			"--command-fd" "0"
  			"--yes" ; overwrite
***************
*** 101,106 ****
--- 102,108 ----
  	  (setq process
  		(apply #'start-process "pgg-gpg" buffer pgg-gpg-program args)))
        (set-default-file-modes orig-mode))
+     (set-process-coding-system process 'utf-8)
      (set-process-filter process #'pgg-gpg-process-filter)
      (set-process-sentinel process #'pgg-gpg-process-sentinel)
      process))
Index: lisp/ChangeLog
===================================================================
RCS file: /usr/local/cvsroot/gnus/lisp/ChangeLog,v
retrieving revision 7.1503
diff -C0 -r7.1503 ChangeLog
*** lisp/ChangeLog	6 Mar 2007 05:16:37 -0000	7.1503
--- lisp/ChangeLog	7 Mar 2007 18:09:28 -0000
***************
*** 0 ****
--- 1,5 ----
+ 2007-03-07  Michaël Cadilhac  <michael@cadilhac.name>
+ 
+ 	* pgg-gpg.el (pgg-gpg-start-process): Ask Gnupg to output everything in
+ 	UTF-8, and treat it as such.
+ 

[-- Attachment #1.3: Type: text/plain, Size: 434 bytes --]


Thanks in advance ! (Well, I have to admit, I didn't like to see
« Michaël » spelled « Michaël » :-))

-- 
 |   Michaël `Micha' Cadilhac       |  Pour les 35-40 ans, l'humour          |
 |   http://michael.cadilhac.name   |       c'est une plus-value.            |
 |   JID/MSN:                       |          -- Guillaume L.               |
 `----  michael.cadilhac@gmail.com  |                                   -  --'

[-- Attachment #2: Type: application/pgp-signature, Size: 188 bytes --]

Re: GPG --status-fd outputs in UTF-8.

[Daiki Ueno]

>>>>> In <87ps7k9anj.fsf@lrde.org> 
>>>>>	michael@cadilhac.name (Michaël Cadilhac) wrote:
>   	  (setq process
>   		(apply #'start-process "pgg-gpg" buffer pgg-gpg-program args)))
>         (set-default-file-modes orig-mode))
> +     (set-process-coding-system process 'utf-8)

At least, this change will cause a problem on decryption, if the
plain text is encoded in non-UTF-8 encoding.

BTW, as you might know, PGG is no longer recommended for No Gnus, and
the current primary backend EasyPG handles status-fd output more
precisely, it's easy to make it decode user-id's which only appear on
status lines.

Regards,
-- 
Daiki Ueno

Re: GPG --status-fd outputs in UTF-8.

[Daiki Ueno]

>>>>> In <526bd3a0-07d1-4009-976f-bf692643d34d@well-done.deisui.org> 
>>>>>	Daiki Ueno <ueno@unixuser.org> wrote:
> >>>>> In <87ps7k9anj.fsf@lrde.org> 
> >>>>>	michael@cadilhac.name (Michaël Cadilhac) wrote:
> >   	  (setq process
> >   		(apply #'start-process "pgg-gpg" buffer pgg-gpg-program args)))
> >         (set-default-file-modes orig-mode))
> > +     (set-process-coding-system process 'utf-8)

> At least, this change will cause a problem on decryption, if the
> plain text is encoded in non-UTF-8 encoding.

s/decryption/encryption/

Regards,
-- 
Daiki Ueno

Re: GPG --status-fd outputs in UTF-8.

[Michaël Cadilhac]

[-- Attachment #1.1: Type: text/plain, Size: 499 bytes --]

Daiki Ueno <ueno@unixuser.org> writes:

>>>>>> In <87ps7k9anj.fsf@lrde.org> 
>>>>>>	michael@cadilhac.name (Michaël Cadilhac) wrote:
>>   	  (setq process
>>   		(apply #'start-process "pgg-gpg" buffer pgg-gpg-program args)))
>>         (set-default-file-modes orig-mode))
>> +     (set-process-coding-system process 'utf-8)
>
> At least, this change will cause a problem on encryption, if the
> plain text is encoded in non-UTF-8 encoding.

You're right, sorry.  Maybe the following?


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1.2: pgg-gpg.patch --]
[-- Type: text/x-patch, Size: 953 bytes --]

--- pgg-gpg.el	02 Mar 2007 00:43:33 +0100	7.27
+++ pgg-gpg.el	12 Mar 2007 18:24:20 +0100	
@@ -120,7 +120,7 @@
 	(beginning-of-line)
 	(while (looking-at ".*\n")	;the input line is finished
 	  (save-excursion
-	    (if (looking-at "\\[GNUPG:] \\([A-Z_]+\\)\\>.*")
+	    (if (looking-at "\\[GNUPG:] \\([A-Z_]+\\)\\> ?\\(.*\\)")
 		(let* ((status (match-string 1))
 		       (symbol (intern-soft (concat "pgg-gpg-status-"
 						    status))))
@@ -128,8 +128,12 @@
 		      (setq pgg-gpg-pending-status-list nil))
 		  (if (and symbol
 			   (fboundp symbol))
-		      (funcall symbol process (buffer-substring
-					       (match-beginning 1)
-					       (match-end 0)))))))
+		      (funcall symbol process
+			       ;; Status value is in UTF-8.
+			       (decode-coding-string
+				(encode-coding-string
+				 (match-string 2)
+				 (car (process-coding-system process)))
+				'utf-8))))))
 	  (forward-line))
 	(setq pgg-gpg-read-point (point)))))

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1.3: egp.patch --]
[-- Type: text/x-patch, Size: 1374 bytes --]

*** epg.el  2007-01-06 00:43:25.000000000 +0100
--- epg.el   2007-03-12 18:14:05.000000000 +0100
***************
*** 1136,1142 ****
                (while (looking-at ".*\n") ;the input line finished
                  (if (looking-at "\\[GNUPG:] \\([A-Z_]+\\) ?\\(.*\\)")
                      (let* ((status (match-string 1))
!                            (string (match-string 2))
                             (symbol (intern-soft (concat "epg--status-"
                                                          status))))
                        (if (member status epg-pending-status-list)
--- 1136,1147 ----
                (while (looking-at ".*\n") ;the input line finished
                  (if (looking-at "\\[GNUPG:] \\([A-Z_]+\\) ?\\(.*\\)")
                      (let* ((status (match-string 1))
!                            ;; Status value is in UTF-8.
!                            (string (decode-coding-string
!                                     (encode-coding-string
!                                      (match-string 2)
!                                      (car (process-coding-system process)))
!                                     'utf-8))
                             (symbol (intern-soft (concat "epg--status-"
                                                          status))))
                        (if (member status epg-pending-status-list)

[-- Attachment #1.4: Type: text/plain, Size: 338 bytes --]


Thanks!

-- 
 |   Michaël `Micha' Cadilhac       |  To be portable,                       |
 |   http://michael.cadilhac.name   |    Just stay on Windows.               |
 |   JID/MSN:                       |                                        |
 `----  michael.cadilhac@gmail.com  |          -- A Microsoft Guy       -  --'

[-- Attachment #2: Type: application/pgp-signature, Size: 188 bytes --]

Re: GPG --status-fd outputs in UTF-8.

[Daiki Ueno]

>>>>> In <873b4ae5hc.fsf@lrde.org> 
>>>>>	michael@cadilhac.name (Michaël Cadilhac) wrote:
> [1.3 egp.patch <text/x-patch (7bit)>]
> *** epg.el  2007-01-06 00:43:25.000000000 +0100
> --- epg.el   2007-03-12 18:14:05.000000000 +0100
> !                            ;; Status value is in UTF-8.
> !                            (string (decode-coding-string
> !                                     (encode-coding-string
> !                                      (match-string 2)
> !                                      (car (process-coding-system process)))
> !                                     'utf-8))

I installed a similar change in CVS.  The next version of EPG will
display your name correctly :-)

Regards,
-- 
Daiki Ueno