Re: saveing/loading symbol table of annymous functions

[David Bateman]

John W. Eaton wrote:
> On 16-May-2007, David Bateman wrote:
>
> | Ok, I checked the patch in (with a minor change so that "make check"
> | runs correctly).
>
> Thanks.  I also made my changes to load_path.  While doing that, I
> noticed
>
>   // Next just search for it anywhere in the
>   // system path
>   string_vector names(3);
>   names(0) = fname + ".oct";
>   names(1) = fname + ".mex";
>   names(2) = fname + ".m";
>
>   dir_path p (load_path::system_path ());
>
>   str = octave_env::make_absolute 
>     (p.find_first_of (names), octave_env::getcwd ())
>
> Why not seaarch the entire load path?  Also, maybe this functionality
> (searching just the system directories) should be an option for the
> load_path::find_X functions?
>
> jwe
>
>   
I deliberately didn't search all of the load path, since if the function
used for the function handle is part of Octave/Matlab it should be found
in Octave/Matlab itself rather than elsewhere. Even if I searched the
loadpath, I'd want the system versions to be taken in precedence over
the other versions in the loadpath. So the order of the loadpath
shouldn't be respected..

Sure why not make this a load_path::find_X function. In that case I
wouldn't need access to load_path::system_path() externally..

Regards
David


-- 
David Bateman                                address@hidden
Motorola Labs - Paris                        +33 1 69 35 48 04 (Ph) 
Parc Les Algorithmes, Commune de St Aubin    +33 6 72 01 06 33 (Mob) 
91193 Gif-Sur-Yvette FRANCE                  +33 1 69 35 77 01 (Fax) 

The information contained in this communication has been classified as: 

[x] General Business Information 
[ ] Motorola Internal Use Only 
[ ] Motorola Confidential Proprietary