[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: shortcut Google search result snooping
- From: Katsumi Yamaoka <yamaoka@xxxxxxx>
- Date: Mon, 16 Apr 2012 08:27:52 +0900
- X-ml-name: emacs-w3m
- X-mail-count: 11799
- References: <87aa2d2h02.fsf@xxxxxxxxxxx>
In [emacs-w3m : No.11798] jidanni@xxxxxxxxxxx wrote:
> Nowadays, e.g.,
> http://www.google.com.tw/search?q=GeoPDF&ie=utf-8&oe=utf-8 just leads to
> a page full of Google snoop links like
> http://www.google.com.tw/url?q=http://en.wikipedia.org/wiki/GeoPDF&sa=U&ei=_meKT5W3IZDSmAXD2uXqCQ&ved=0CBYQFjAA&usg=AFQjCNHWLQW_0J920pZEjdztKO-oTTTGgA
> Maybe there should be some functionality that automatically rewrites
> this back to a direct call to http://en.wikipedia.org/wiki/GeoPDF .
> I don't think letting Google know our User-Agent is worth the extra connection.
That's easy to do using w3m-filter:
(setq w3m-use-filter t)
(require 'w3m-filter)
(defun w3m-filter-google (url)
"Extract real urls in Google search."
(goto-char (point-min))
(while (re-search-forward "\\(<a[\t\n ]+href=\"\\)/url\\?q=\\([^&]+\\)"
nil t)
(replace-match "\\1\\2\"")))
This replaces the default Google filter function that w3m-filter.el
has been providing thus far, but I think this becomes the new default.
Does anyone know this causes any disadvantage to users?