ffap vs. URLs in UTF-8

[jidanni@xxxxxxxxxxx]

Just curious, how do you emacs-w3m pros "click" on these links that are
not linked on the following page. E.g.,
$ w3m -dump -cols 999 http://pots.tw/node/4415|perl -nwle '/\S(http\S+)/&&print $1'
http://tw.myblog.yahoo.com/taipeigayinn/
http://coswas.org/01coswas/966
http://www.psygarden.com.tw/images/activities/MjAxMDAyMDgxMzE2NDI=/MjAxMDAyMDgxMzE2NDI=.03redingclub-signup.doc
http://spreadsheets.google.com/viewform?formkey=dGEzVG95Vl9Mc3lCZWVVdG5jem0zX2c6MA
http://tinyurl.com/ybfjulc
http://0rz.tw/XGtTg下載報名表，填寫完後email到CLPENG@xxxxxxxxxxxxxx，報名時間至3/21截止。
http://woman1.pixnet.net/blog)
http://spreadsheets.google.com/viewform?formkey=dE5IYWtzSlQwUl9kMGZEMVRULXhDbUE6MA
http://msmtaiwan.blogspot.
http://www.wretch.cc/blog/moongirl1225/21391679
http://www.wretch.cc/blog/taboo126/21681160

ffap fails to detect them because its matcher is too strict, so C-x C-f
won't work. One ends up needing to use C-k, then C-x C-f ... C-y. Yuck.

My point of this message is I hope someone will change emacs' ffap
regexp, as posting to bug-gnu-emacs didn't get any change ever.