[Bioperl-l] WrapperBase / StandAloneBlast executable() method confusion

[Chris Fields]

I'm not sure, but you might want to bring Torsten in on this as he  
took over maintaining StandAloneBlast.  Much of the confusion may  
stem from the independent evolution of StandAloneBlast and WrapperBase.

Also, (a bit unrelated), there were plans for unifying the  
Bio::Tools::Run BLAST modules described here:

http://www.bioperl.org/wiki/Module:Bio::Tools::Run::RemoteBlast

Seemed like there was a general consensus at the time on the need to  
refactor StandAloneBlast and RemoteBlast code, so maybe the best  
place to start is StandAloneBlast (the others could be added in from  
there).  We could just deprecate use of the older modules at some  
point in favor of the new scheme.

chris

On Apr 24, 2007, at 5:21 AM, Sendu Bala wrote:

> Hi,
>
> I'm a little unsure of the intent for executable() in wrapper modules.
> The WrapperBase version of the method and the StandAloneBlast version
> have the same POD but different implementations.
>
> WrapperBase takes as a first arg an 'exe' which it will blindly  
> trust is
> the path to a working executable. (That doesn't seem sensible  
> already.)
> It is only capable of storing one such path.
>
> If no arg is supplied it uses program_path() (which uses  
> program_name())
> to find the executable. Failing that it does a further direct test on
> program_name() to see if its executable.
>
>
> StandAloneBlast takes as a first arg merely the name of your exe and
> also (undocumented) the path to the corresponding executable (which is
> tested to see if it really executable). It can store executable paths
> for multiple different exenames (corresponding better with the docs  
> for
> the first arg: "name of executable to set path to").
>
> If no second arg is supplied it does something similar to WrapperBase,
> except that it uses the first arg exename (or a default if that wasn't
> supplied) in place of program_name().
>
>
> I'm trying to generalize this so StandAloneBlast can just use the
> WrapperBase version (and so other wrappers can then store executable
> paths for different sub-programs). Any suggestions for a good way of
> melding these two together whilst somehow retaining backward  
> compatibility?
> _______________________________________________
> Bioperl-l mailing list
> Bioperl-l at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/bioperl-l

Christopher Fields
Postdoctoral Researcher
Lab of Dr. Robert Switzer
Dept of Biochemistry
University of Illinois Urbana-Champaign