Re: Czech characters displayed incorrectly

[Katsumi Yamaoka <yamaoka@xxxxxxx>]

>>>>> In [emacs-w3m : No.09975] niels.giesen@xxxxxxxxx wrote:

> Problem with czech characters in w3m. For instance, the sequence říň
> (if messed up, this means r with a haček, long i and n with a haček)
> is displayed incorrectly.

Well, it might be irrelevant to your problem but we've been aware
of the cookie problem in emacs-w3m.  For instance, www.google.cz
doesn't return an expected answer for searching of the word "říň"
even if having customized the preferred language into česky[1].
A possible workaround is to add the following one to the
~/.emacs-w3m.el file in order to enable emacs-w3m to accept
cookies.  It is effective only with .cz sites that behave like
Google.cz, though.

(eval-after-load "w3m-cookie"
  '(let ((adds '("cz"))
	 (tlds (delete "" (split-string
			   w3m-cookie-two-dot-domains-regexp
			   "[^a-z]+"))))
     (mapc (lambda (add) (setq tlds (delete add tlds))) adds)
     (setq w3m-cookie-two-dot-domains-regexp
	   (concat
	    "\\.\\(?:"
	    (mapconcat 'identity (nconc tlds adds) "\\|")
	    "\\)$"))))

[1] It is necessary at least for me since Google.cz prefers
Japanese text by default for the client that accesses from Japan.

> When I retrieve the page with wget, and then visited as a file, it
> gets displayed right. So this is the expected behaviour:

Could you let me know the url(s) in question?  I'll look into how
emacs-w3m works with it.

> I checked the same in stand-alone w3m, which displayed everything
> correctly.  Please note these are not the only czech characters
> displayed incorrectly.  If you like me to provide the whole
> alphabet, I shall.  One strange thing I saw with some other
> characters is that they seemed to be displayed as Thai characters.
> Please let me know if you want more info.

[...]

> emacs-w3m-version
>  => "1.4.4"
> emacs-version
>  => "GNU Emacs 22.1.1 (i486-pc-linux-gnu, X toolkit, Xaw3d scroll bars)\n of 2007-11-06 on terranova, modified by Ubuntu"

[...]

> w3m-version
>  => "w3m/0.3.2+mee-p24-19+moe-1.5.0"
> w3m-type
>  => w3mmee

Emacs-w3m v1.4.4 is a bit old, and I don't have w3mmee.  So, I'll
try it using the latest emacs-w3m CVS and the latest w3m CVS first.

Regards,