On Wednesday 30 May 2001 14:58, Stephan Kulow wrote:
> Hi!
> 
> I have a big problem with KURL and encodings. At first this is the output of 
> kurltest when setting russian:
> 
> kurltest: locale8bit : checking 'file:/home/dfaure/konqtests/Matériel' 
> against expected value 'file:/home/dfaure/konqtests/Matériel'... ok
> kurltest: pretty : checking 'file:/home/dfaure/konqtests/Mat�iel' against 
> expected value 'file:/home/dfaure/konqtests/Matériel'... KO !

Hehe, I recognize this test.
Well, I don't think it makes any sense in Russian, in the first place. 
This is a latin1 test. The "expected value" differs with the locale encoding, obviously.
(And especially prettyURL isn't meant to give the same result in all encodings,
on the contrary).

> And above that I get this with some added debug output:
> libkonq: KonqOperations::rename newPath=/home/coolo/афыв7
> kio_file: rename file:/home/coolo/%3F%3F%3F%3F -> 
> file:/home/coolo/%3F%3F%3F%3F7
>
> I don't know how this appears in your mailer, but in the first line between 
> /home/coolo and 7 are four russian characters. These end up as %3F in the URL 
> kio_file gets. The encoding (or kodipowka as the russian call it :) i s 
> iso10464-1. 

Can you look into the [file-]static "encode" method of KURL ?
Since you're not using drag-n-drop, encoding_hint should always be 0... so this
uses local8Bit() to convert to QCString, and from there the chars are simply encoded
by value. So if you get %3F, it means that either the conversion ruins the string, or
it's whatever is passed to encode() that is already wrong.....

-- 
David FAURE, david@mandrakesoft.com, faure@kde.org
http://perso.mandrakesoft.com/~david/, http://www.konqueror.org/
KDE, Making The Future of Computing Available Today