ValueCD WebMaster wrote:

> [snip]
>
> > I see a "font class=name>Pege/Wilfer/Pauer/Pege"
> > line and a line with an html
> > link that has the encoded %A5s in the href and
> > the unencoded
> > "Pege/Wilfer/Pauer/Pege" as the displayed tag text.
>
> Actually, the server has sent you a wrong page
> (a substitute).
> It understood only the first word from the URL
> ("Pege") and has sent you a result page WITH
> A LINK to the correct page instead of the correct
> page itself - which is exactly the link above,
> the one I have had the problem with.
>

Actually, the link it sent back is *exactly* the same as the one I was
trying to get. So, if I extracted the link from the page and fetched
that one instead, I'd get the exact same page back a second time.

> So it looks like even Linux has a problem with
> multiple "%" statements (I am on Win32 platform).

I'm not so sure about that. Tcpdump shows that the outgoing request
looks like this:

GET http://www.musicline.de/de/artist/Pege%A5Wilfer%A5Pauer%A5Pege
HTTP/1.1

with multiple %A5's in place and correct. Never mind, as long as it's
now working, that's all good.

Ralph

> All of the above notwithstanding, thank you for
> responding, Ralph, I appreciate it.
>
> And of course thanks to Daniel; he actually solved
> my problem by writing "Your shell or something
> eats the % letters" - it turned out I had to
> DOUBLE the % characters to get it right.
>
> So the correctly encoded URL is
> http://www.musicline.de/de/artist/Pege%%A5Wilfer%%A5Pauer%%A
> 5Pege
> (= each %-character doubled)
> and it brings the correct page back from the server.
>
> Thanks again for your help!!

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
Received on 2002-08-29