[Bioperl-l] Use of Root.pm versus RootI.pm

Chris Fields cjfields at uiuc.edu
Tue Oct 3 13:49:35 EDT 2006


Perl6 already has added flexibility for separation of
implementation/interface (I believe they are called roles).  

http://dev.perl.org/perl6/doc/design/syn/S12.html

To tell the truth, I'm not sure about Perl 5, except the way the Bioperl
devs have up the distinction between interface and implementation.  However,
I find the way we use interfaces is very simple (set up interface with
some/all methods as unimplemented, use the module as an abstract base class,
then override the unimplemented methods).  It works for me.

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

> -----Original Message-----
> From: Stephen Gordon Lenk [mailto:slenk at emich.edu]
> Sent: Tuesday, October 03, 2006 11:45 AM
> To: Chris Fields
> Cc: 'Hilmar Lapp'; 'Torsten Seemann'; 'bioperl-l'
> Subject: Re: [Bioperl-l] Use of Root.pm versus RootI.pm
> 
> The separation of interface and implementation is generally
> regarded as a good idea. Right now the Bioperl community is
> doing this as part of the implementation of Bioperl. I suggest
> that this is an example of something which you might want to
> have as part of the Perl implementation. If Perl 6 (or even
> Perl 5) does not have this as a core part of the language or
> as a standard package (reusable by all in a common fashion),
> you may want to suggest to the Perl implementers that a way
> for interface/implementation distinctions be made part of the
> core language. My 2 cents, as you people are the experts on
> your own code.
> 
> 
> ----- Original Message -----
> From: Chris Fields <cjfields at uiuc.edu>
> Date: Tuesday, October 3, 2006 10:29 am
> Subject: Re: [Bioperl-l] Use of Root.pm versus RootI.pm
> 
> > > The interface classes (those ending in 'I') should actually inherit
> > > from RootI, not Root.
> > >
> > > In reality this recommendation is more theoretical than it makes
> > that> much of a difference I think. The motivation is that interface
> > > classes should not determine the actual implementation of a class
> > > (hash ref, array ref, whatever), and since Root.pm contains lots of
> > > implementation using a hash ref that decision will basically have
> > > been made.
> > >
> > > On the contrary though, RootI contains implementation too, although
> > > I'm not sure it would prescribe the object implementation as
> opposed
> > > to merely implementing static methods (like throw(), warn(), etc).
> > > That would need to be checked.
> > >
> > > 	-hilmar
> >
> > The constructor in Bio::Root::RootI lets one know that its use is
> > deprecated, so you shouldn't have any cases of 'our
> > qw(Bio::Root::RootI)';there should be some way of inheriting Root
> > directly or indirectly.  I would
> > say that any direct use of RootI is not good practice, though.
> > For the
> > current implementation we should only inherit Bio::Root::Root, which
> > implements RootI.
> >
> > Is there any reason to shut off the warning with BIOPERLDEBUG?
> >
> > >From RootI:
> >
> > sub new {
> >  my $class = shift;
> >  my @args = @_;
> >  unless ( $ENV{'BIOPERLDEBUG'} ) {
> >      carp("Use of new in Bio::Root::RootI is deprecated.  Please use
> > Bio::Root::Root instead");
> >  }
> >  eval "require Bio::Root::Root";
> >  return Bio::Root::Root->new(@args);
> > }
> >
> >
> > Christopher Fields
> > Postdoctoral Researcher - Switzer Lab
> > Dept. of Biochemistry
> > University of Illinois Urbana-Champaign
> >
> >
> >
> > >
> > > On Oct 3, 2006, at 12:30 AM, Torsten Seemann wrote:
> > >
> > > > My understanding is that all Bioperl-compliant classes should
> > inherit> > from Bio::Root::Root, not Bio::Root::RootI.
> > > >
> > > > Additionally, if functions such as throw() or _rearrange() are
> > to be
> > > > used without a class instance reference, they are to be used
> > as class
> > > > methods via Bio::Root::Root, not Bio::Root::RootI.
> > > >
> > > > Is this correct?
> > > >
> > > > My naive audit of bioperl-live CVS brought up the following
> > > > statistics:
> > > >
> > > > # Root.pm
> > > > /cvs/bioperl-live $ grep -r 'use Bio::Root::Root;' Bio | wc -l
> > > > 26
> > > > /cvs/bioperl-live $ grep -r 'use base.*Bio::Root::Root' Bio |
> > wc -l
> > > > 346
> > > >
> > > > # RootI.pm
> > > > /cvs/bioperl-live $ grep -r 'use Bio::Root::RootI;' Bio | wc -l
> > > > 9
> > > > /cvs/bioperl-live $ grep -r 'use base.*Bio::Root::RootI' Bio |
> > wc -l
> > > > 79
> > > >
> > > > My guess would be that all RootI should be changed to plain
> > Root ?
> > > >
> > > > Any help appreciated,
> > > >
> > > > --
> > > > Dr Torsten Seemann               http://www.vicbioinformatics.com
> > > > Victorian Bioinformatics Consortium, Monash University, Australia
> > > >
> > > > _______________________________________________
> > > > Bioperl-l mailing list
> > > > Bioperl-l at lists.open-bio.org
> > > > http://lists.open-bio.org/mailman/listinfo/bioperl-l
> > > >
> > >
> > > --
> > > ===========================================================
> > > : Hilmar Lapp  -:-  Durham, NC  -:-  hlapp at gmx dot net :
> > > ===========================================================
> > >
> > >
> > >
> > >
> > >
> > > _______________________________________________
> > > Bioperl-l mailing list
> > > Bioperl-l at lists.open-bio.org
> > > http://lists.open-bio.org/mailman/listinfo/bioperl-l
> >
> > _______________________________________________
> > Bioperl-l mailing list
> > Bioperl-l at lists.open-bio.org
> > http://lists.open-bio.org/mailman/listinfo/bioperl-l
> >



More information about the Bioperl-l mailing list