Re: Find-file at point cannot manage tramp file syntax

[Stéphane Senesi]

Michael Albinus wrote, On 05/08/2012 13:16:
> Stéphane Senesi<address@hidden>  writes:
>
>    
> > Hello
> >      
> Hi,
>
>    
> > I am using emacs 23.1.1 with tramp 2.2.3
> >
> > When working on a remote file, which tramp location looks like
> > "rsh:address@hidden:...", and invoking find-file-at-point at a place which
> >      
> I guess you mean "/rsh:address@hidden:...".
>    
Yes, you're right
>    
> > shows a valid file reference on the remote host, I got the message :
> > "rsh : Unknown host" in a buffer named "*tramp/rsh address@hidden" (while
> > the relevant buffer, which do exist, is "*tramp/rsh address@hidden")
> >
> > This does not happen in dired-mode (maybe find-file-at-point is here
> > re-mapped to find-file or dired-find-file ?)
> >
> > It looks like file-find-at-point should, and does not, discard the
> > "rsh:" string before invoking tramp to get the file. Is it a
> > compatibility issue (between emacs and tramp versions) ?
> >      
> I cannot reproduce it here with Emacs 23.3.1. Could you, please, check
> whether this happens also with "emacs -Q"? If yes, I would need a
> precise recipe for reproducing.
>    

I investigated further. The problem only appears when the path close to 
the point is either an absolute path or a path with a tilde ~ :
    - using 'emacs -Q', ffap erroneously (?) tries to find the file on 
the local machine, and hence, for the case of an absolute path, proposes 
to truncate the file path to the longest one which matches the local 
file system
    - using my settings, I get the behaviour described above

When discarding from my settings the "(require 'ange-ftp)" that was 
intended to allow for :
    (setq ange-ftp-skip-msgs (concat ange-ftp-skip-msgs "\\|^500 This 
security scheme is not implemented"))
I am back to the 'emacs -Q' behaviour

So, I have only one question left : could/should tramp+ffap interpret 
absolute and tilde path found at the point of a remote file as remote 
paths ?

Anyway, thanks for the hint on the method.

Stéphane

>    
> > Best regards
> >      
> Best regards, Michael.
>
>    


--
Stéphane Sénési
Ingénieur - équipe Assemblage du Système Terre
Centre National de Recherches Météorologiques
Groupe de Météorologie à Grande Echelle et Climat

CNRM/GMGEC/ASTER
42 Av Coriolis
F-31057 Toulouse Cedex 1

+33.5.61.07.99.31 (Fax :....9610)