Re: Emacs-W3m bug

[David Hansen <david.hansen@xxxxxxx>]

On Sun, 4 May 2008 10:10:09 +0800 Lazycat Manatee wrote:

> Hi.
> I'm use Emacs-W3m (1.5) in Emacs (23.0.60.2).
> Blow is my function about search in Google Blog.
>
> (defun w3m-search-google-blog (what)
>   "Use Google blog search for WHAT."
>   (interactive "sGoogle Blog: ")
>   (w3m-goto-url-new-session (concat
> "http://blogsearch.google.cn/blogsearch?hl=zh-CN&ie=UTF-8&q=";
> 				    (w3m-url-encode-string what 'utf-8))))

Not your that it solves your problem, but there is a more elegant way to
add new search engines to w3m:

(add-to-list 'w3m-search-engine-alist
             '("google-blog"
               "http://blogsearch.google.cn/blogsearch?hl=zh-CN&ie=UTF-8&q=%s";
               utf-8))

This way you can use your favorite search engine with C-u s / C-u S.

> If I use "w3m-search-google-blog" search "奥运会",
> so the coding of first page of Google search result is right(ie=UTF-8),
> but the coding of second page of Google search result is wrong
> (automatic change to ie=GB2312).

Isn't it googles fault? It's their server which chooses the encoding of
the reply.

A workaround may be using google.com instead of google.cn.  This may
also give (depending on the search topic) more interesting results.

David