[Bioperl-l] Auto-method caller proposal

[Chris Fields]

On Jan 9, 2007, at 9:12 AM, Sendu Bala wrote:

> Nathan (Nat) Goodman wrote:
>>>> On Jan 3, 2007, at 1:09 PM, Sendu Bala wrote:
>>
>>> Alternatively, would the system work if only Bio::Root::RootI was
>>> based on Class::AutoClass? What would the necessary code be for
>>> changes in RootI and then changes in an existing run-wrapper for
>>> example?
>>
>> Yes, it should work for Bio::Root::RootI to inherit from AutoClass.
>> This, however, would impose AutoClass on every BioPerl-er whether  
>> they
>> want it or not :)
>
> As a generally useful thing I'd like to see all Bioperl modules have
> easy access to this functionality, just as they can currently call
> _rearrange(). So yes, we would impose AutoClass on everyone. This may
> not be a major burden since it is already an optional pre- 
> requisite. So,
> before I investigate using Class::AutoClass instead of my own proposed
> method, does anyone feel there are good benefits of Class::AutoClass
> over my proposed method (considering I already added a -synonym option
> to it), and do those benefits outweigh having Class::AutoClass as a
> Bioperl installation requirement?

To me, this goes back to Aaron's argument about 'reinventing the  
wheel.'  If you can accomplish what you need using Class::AutoClass,  
why not use it?  Or as Nat suggests, use the relevant  
Class::AutoClass code directly in RootI.  However, if you believe  
your method is simpler, by all means use that.

The main criticisms about using autogenerated methods (summarized by  
Nat) seem to be that (1) they cause problems with understanding code  
and debugging, and (2) too magic argument processing.  However, there  
are many instances where AUTOLOAD is being used in bioperl to deal  
with general get/setters, or a similar kludge (like the heredoc eval 
{} kludge I posted earlier) is used instead.  Based on that and your  
previous post there is obviously a need for some way to autogenerate  
methods like get/setters, even if that need isn't expressed openly.

chris