Re: [Html2wml] Strange page doesn't work

[<mad...@fr...>]

On mardi, f=E9v 11, 2003, at 18:56 Europe/Paris, fri...@gm... =
wrote:

> Salut S=E9bastien,
>
> there's still a problem with the URL because it has a question mark in =
=20
> it; I think the
> html2wml.cgi thinks that there's a new argument for it. But instead, =20=

> everything that
> follows is still part of the URL to be retrieved. That's why I get the =
=20
> "HTML Error" back
> from the server, because it is given no arguments, i.e.:
>
> http://bahnhofstafel.db-
> ris.de/bht/internet/=20
> tafel.html?AnAb=3DAB&m=3Ddn&ident=3D&seqnr=3D&L=3D&bhf=3D8002548
>
> would become
>
> http://bahnhofstafel.db-ris.de/bht/internet/tafel.html
>
> which is not enough.

I loaded this URL and everything seems to work fine. I also tried with =20=

the result of a Google search and it also worked. How are you using =20
Html2Wml? as a CGI? Which browser are you using?

> The same page with a different URL that I created works fine:
> http://shorl.com/bistosajemuby
>
> Do you know if there's a way to escape the question mark? I've tried =20=

> putting \? instead
> or enclosing the url in quotation marks, but that doesn't work...

The question marks are already escaped using the URL syntax (%xx), but =20=

the browser can decode them and decide they're errors. An heavy =20
workaround is to encode all the URLs in Base64 (that's what I was =20
testing this week-end). I will try to integrate this but it requires =20
more work.

Regards,

S=E9bastien Aperghis-Tramoni
  -- - --- -- - -- - --- -- - --- -- - --[ http://maddingue.org ]=