On Mon, 21 May 2012, Manfred Schwarb wrote:

> This URL was gathered with "grep" from a preceding html page, so this means
> the web programmer put an illegal URL inside the <a href=...> token.
>
> This programmer obviously assumed the web client does the needed sanitizing
> (which, it seems, most of them do). The programmer of course tested his web
> page only with web browsers, which seem to correct this miscoding.

That is unfortunately how the web browser world works, but I strongly disagree
with that view. They keep obeying to all sorts of mishaps users do only
because another browser does it.

When it comes to URL parsing and encoding, curl might one day have to adopt to
a more browser-like behavior but I'm not going there willingly.

> So I have to do the cleanup myself replacing spaces with %20.

Yes, I believe that's the annoying reality. :-/

-- 
  / daniel.haxx.se
-------------------------------------------------------------------
List admin: http://cool.haxx.se/list/listinfo/curl-users
FAQ:        http://curl.haxx.se/docs/faq.html
Etiquette:  http://curl.haxx.se/mail/etiquette.html