Re: Can someone please comment on this short program

[Matthias Kaeppler <matthias finitestate org>]

Murray Cumming wrote:
> I probably heard somewhere that all URIs are UTF-8 encoded. I didn't take
> the time to look at the appropriate W3C specification, RFC, etc. I also
> assumed that URIs are escaped.
>
> In your test case, you are using the C++ wrapper for
> gnome_vfs_get_uri_from_local_path(filepath).
> I wonder what this does? Does it just add file:// at the start without
> doing any escaping/encoding-conversoin?
>
> http://developer.gnome.org/doc/API/gnome-vfs/gnome-vfs-gnome-vfs-utils.html#GNOME-VFS-GET-URI-FROM-LOCAL-PATH

I don't know, the API reference doesn't say anything about conversion so 
I must assume no conversion is happening. I don't think this function is 
relevant to this issue anyway, though I may be wrong.

> > (that's because create() takes a Glib::ustring, and if
> > you pass an std::string in place of a Glib::ustring--such as the one
> > returned from FileInfo::get_name()--a conversion attempt will be made).
>
>
> I don't think that's true, but maybe I'm looking at the wrong part of the
> Glib::ustring string source. The ustring(std::string) constructor does a
> simple assignment.

Yes you're right; my excuses. Probably the conversion happens somewhere 
inside the create() function or in a function called within create() and 
not in any ustring ctor.
In any way, calling create() with a string that is not UTF-8 encoded 
raises an exception on my system (unless it only contains the first 127 
ASCII characters).

Example:
	RefPtr<Uri> uri = Uri::create("täst");

matthias:testing$ ./a.out
terminate called after throwing an instance of 'Glib::ConvertError'
Uri: Aborted

Please note at this point, that I could have as well passed an 
std::string encoded in something other than UTF-8 to create(), and not 
just some hardwired string which happens to not be encoded in my 
locale's encoding due to my text editor being ignorant (in your face, 
Emacs!).

The vicious circle here is that by converting the argument safely to 
UTF-8 to let create() succeed, I thereby break functions like uri_exists().

Regards,
Matthias