Re: wrong -O arg to w3m-m17n

[ari@xxxxxxxxxxxxxxxx (ARISAWA Akihiro)]

Hi,

>>>>> In [emacs-w3m : No.05551] 
>>>>>	कपिल <klchxbec@m-net.arbornet.org> wrote:

> I'm trying to use emacs-w3m-1.3.6 with w3m-m17n-0.4.1 on
> emacs-21.3 to view utf-8 encoded devanagari (hindi) pages
> like in http://www.bbc.co.uk/hindi/

> This resulted in emacs-w3m displaying ? for utf characters. I

[...]

> Can someone fix emacs-w3m to detect the presence of utf on
> emacs-21.3? I also saw that there is no 'un-define on
> emacs-21.3.

Thank you for your report.
I had committed the attached changes. If you are not using Mule-UCS,
please change w3m-output-coding-system to 'utf-8. If you are using
Mule-UCS, there is no need to change w3m-output-coding-system or
w3m-halfdump-command-arguments.

;; Emacs21 built-in utf-8 cannot treat so many characters. 
;; For that reason, I don't want to use it by default.

Regards,
ARISAWA Akihiro
Index: w3m.el
===================================================================
RCS file: /storage/cvsroot/emacs-w3m/w3m.el,v
retrieving revision 1.831
diff -u -r1.831 w3m.el
--- w3m.el	17 Jul 2003 14:52:36 -0000	1.831
+++ w3m.el	24 Jul 2003 12:02:35 -0000
@@ -426,7 +426,8 @@
   (cond
    ((not (featurep 'mule)) 'iso-8859-1)
    ((eq w3m-type 'w3mmee) 'ctext)
-   ((eq w3m-type 'w3m-m17n) (if w3m-use-mule-ucs 'utf-8 'iso-2022-7bit-ss2))
+   ((eq w3m-type 'w3m-m17n)
+    (if (featurep 'un-define) 'utf-8 'iso-2022-7bit-ss2))
    (w3m-accept-japanese-characters 'w3m-euc-japan)
    (t 'w3m-iso-latin-1))
   "*Coding system for read operations of `w3m'."
@@ -1764,7 +1765,9 @@
 	       "-o" "strict_iso2022=0"
 	       '(if charset "-I") 'charset
 	       "-O"
-	       (if w3m-use-mule-ucs "UTF-8" "ISO-2022-JP-2" )))
+	       (if (eq w3m-output-coding-system 'utf-8)
+		   "UTF-8"
+		 "ISO-2022-JP-2" )))
 	((eq w3m-input-coding-system 'w3m-euc-japan)
 	 (list "-halfdump" "-I" "e"))
 	(t (list "-halfdump")))